You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@spark.apache.org by Tianhua huang <hu...@gmail.com> on 2019/09/12 03:06:38 UTC
Re: Ask for ARM CI for spark
Hi all,
For the whole work process of spark ARM CI, we want to make 2 things clear.
The first thing is:
About spark ARM CI, now we have two periodic jobs, one job[1] based on
commit[2](which already fixed the replay tests failed issue[3], we made a
new test branch based on date 09-09-2019), the other job[4] based on spark
master.
The first job we test on the specified branch to prove that our ARM CI is
good and stable.
The second job checks spark master every day, then we can find whether the
latest commits affect the ARM CI. According to the build history and
result, it shows that some problems are easier to find on ARM like
SPARK-28770 <https://issues.apache.org/jira/browse/SPARK-28770>, and it
also shows that we would make efforts to trace and figure them out, till
now we have found and fixed several problems[5][6][7], thanks everyone of
the community :). And we believe that ARM CI is very necessary, right?
The second thing is:
We plan to run the jobs for a period of time, and you can see the result
and logs from 'build history' of the jobs console, if everything goes well
for one or two weeks could community accept the ARM CI? or how long the
periodic jobs to run then our community could have enough confidence to
accept the ARM CI? As you suggested before, it's good to integrate ARM CI
to amplab jenkins, we agree that and we can donate the ARM instances and
then maintain the ARM-related test jobs together with community, any
thoughts?
Thank you all!
[1]
http://status.openlabtesting.org/job/spark-unchanged-branch-unit-test-hadoop-2.7-arm64
[2]
https://github.com/apache/spark/commit/0ed9fae45769d4b06b8cf8128f462f09ff3d9a72
[3] https://issues.apache.org/jira/browse/SPARK-28770
[4]
http://status.openlabtesting.org/builds?job_name=spark-master-unit-test-hadoop-2.7-arm64
[5] https://github.com/apache/spark/pull/25186
[6] https://github.com/apache/spark/pull/25279
[7] https://github.com/apache/spark/pull/25673
On Fri, Aug 16, 2019 at 11:24 PM Sean Owen <sr...@gmail.com> wrote:
> Yes, I think it's just local caching. After you run the build you should
> find lots of stuff cached at ~/.m2/repository and it won't download every
> time.
>
> On Fri, Aug 16, 2019 at 3:01 AM bo zhaobo <bz...@gmail.com>
> wrote:
>
>> Hi Sean,
>> Thanks for reply. And very apologize for making you confused.
>> I know the dependencies will be downloaded from SBT or Maven. But the
>> Spark QA job also exec "mvn clean package", why the log didn't print
>> "downloading some jar from Maven central [1] and build very fast. Is the
>> reason that Spark Jenkins build the Spark jars in the physical machiines
>> and won't destrory the test env after job is finished? Then the other job
>> build Spark will get the dependencies jar from the local cached, as the
>> previous jobs exec "mvn package", those dependencies had been downloaded
>> already on local worker machine. Am I right? Is that the reason the job
>> log[1] didn't print any downloading information from Maven Central?
>>
>> Thank you very much.
>>
>> [1]
>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/job/spark-master-test-maven-hadoop-2.6-ubuntu-testing/lastBuild/consoleFull
>>
>>
>> Best regards
>>
>> ZhaoBo
>>
>> [image: Mailtrack]
>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>> notified by
>> Mailtrack
>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/08/16
>> 下午03:58:53
>>
>> Sean Owen <sr...@gmail.com> 于2019年8月16日周五 上午10:38写道：
>>
>>> I'm not sure what you mean. The dependencies are downloaded by SBT and
>>> Maven like in any other project, and nothing about it is specific to Spark.
>>> The worker machines cache artifacts that are downloaded from these, but
>>> this is a function of Maven and SBT, not Spark. You may find that the
>>> initial download takes a long time.
>>>
>>> On Thu, Aug 15, 2019 at 9:02 PM bo zhaobo <bz...@gmail.com>
>>> wrote:
>>>
>>>> Hi Sean,
>>>>
>>>> Thanks very much for pointing out the roadmap. ;-). Then I think we
>>>> will continue to focus on our test environment.
>>>>
>>>> For the networking problems, I mean that we can access Maven Central,
>>>> and jobs cloud download the required jar package with a high network speed.
>>>> What we want to know is that, why the Spark QA test jobs[1] log shows the
>>>> job script/maven build seem don't download the jar packages? Could you tell
>>>> us the reason about that? Thank you. The reason we raise the "networking
>>>> problems" is that we found a phenomenon during we test, if we execute "mvn
>>>> clean package" in a new test environment(As in our test environment, we
>>>> will destory the test VMs after the job is finish), maven will download the
>>>> dependency jar packages from Maven Central, but in this job
>>>> "spark-master-test-maven-hadoop" [2], from the log, we didn't found it
>>>> download any jar packages, what the reason about that?
>>>> Also we build the Spark jar with downloading dependencies from Maven
>>>> Central, it will cost mostly 1 hour. And we found [2] just cost 10min. But
>>>> if we run "mvn package" in a VM which already exec "mvn package" before, it
>>>> just cost 14min, looks very closer with [2]. So we suspect that downloading
>>>> the Jar packages cost so much time. For the goad of ARM CI, we expect the
>>>> performance of NEW ARM CI could be closer with existing X86 CI, then users
>>>> could accept it eaiser.
>>>>
>>>> [1] https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/
>>>> [2]
>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/job/spark-master-test-maven-hadoop-2.6-ubuntu-testing/lastBuild/consoleFull
>>>>
>>>> Best regards
>>>>
>>>> ZhaoBo
>>>>
>>>>
>>>>
>>>>
>>>> [image: Mailtrack]
>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>> notified by
>>>> Mailtrack
>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/08/16
>>>> 上午09:48:43
>>>>
>>>> Sean Owen <sr...@gmail.com> 于2019年8月15日周四 下午9:58写道：
>>>>
>>>>> I think the right goal is to fix the remaining issues first. If we set
>>>>> up CI/CD it will only tell us there are still some test failures. If it's
>>>>> stable, and not hard to add to the existing CI/CD, yes it could be done
>>>>> automatically later. You can continue to test on ARM independently for now.
>>>>>
>>>>> It sounds indeed like there are some networking problems in the test
>>>>> system if you're not able to download from Maven Central. That rarely takes
>>>>> significant time, and there aren't project-specific mirrors here. You might
>>>>> be able to point at a closer public mirror, depending on where you are.
>>>>>
>>>>> On Thu, Aug 15, 2019 at 5:43 AM Tianhua huang <
>>>>> huangtianhua223@gmail.com> wrote:
>>>>>
>>>>>> Hi all,
>>>>>>
>>>>>> I want to discuss spark ARM CI again, we took some tests on arm
>>>>>> instance based on master and the job includes
>>>>>> https://github.com/theopenlab/spark/pull/13 and k8s integration
>>>>>> https://github.com/theopenlab/spark/pull/17/ , there are several
>>>>>> things I want to talk about:
>>>>>>
>>>>>> First, about the failed tests:
>>>>>> 1.we have fixed some problems like
>>>>>> https://github.com/apache/spark/pull/25186 and
>>>>>> https://github.com/apache/spark/pull/25279, thanks sean owen and
>>>>>> others to help us.
>>>>>> 2.we tried k8s integration test on arm, and met an error: apk
>>>>>> fetch hangs, the tests passed after adding '--network host' option for
>>>>>> command `docker build`, see:
>>>>>>
>>>>>> https://github.com/theopenlab/spark/pull/17/files#diff-5b731b14068240d63a93c393f6f9b1e8R176
>>>>>> , the solution refers to
>>>>>> https://github.com/gliderlabs/docker-alpine/issues/307 and I don't
>>>>>> know whether it happened once in community CI, or maybe we should submit a
>>>>>> pr to pass '--network host' when `docker build`?
>>>>>> 3.we found there are two tests failed after the commit
>>>>>> https://github.com/apache/spark/pull/23767 :
>>>>>> ReplayListenerSuite:
>>>>>> - ...
>>>>>> - End-to-end replay *** FAILED ***
>>>>>> "[driver]" did not equal "[1]" (JsonProtocolSuite.scala:622)
>>>>>> - End-to-end replay with compression *** FAILED ***
>>>>>> "[driver]" did not equal "[1]" (JsonProtocolSuite.scala:622)
>>>>>>
>>>>>> we tried to revert the commit and then the tests passed, the
>>>>>> patch is too big and so sorry we can't find the reason till now, if you are
>>>>>> interesting please try it, and it will be very appreciate if
>>>>>> someone can help us to figure it out.
>>>>>>
>>>>>> Second, about the test time, we increased the flavor of arm instance
>>>>>> to 16U16G, but seems there was no significant improvement, the k8s
>>>>>> integration test took about one and a half hours, and the QA test(like
>>>>>> spark-master-test-maven-hadoop-2.7 community jenkins job) took about
>>>>>> seventeen hours(it is too long :(), we suspect that the reason is the
>>>>>> performance and network,
>>>>>> we split the jobs based on projects such as sql, core and so on, the
>>>>>> time can be decrease to about seven hours, see
>>>>>> https://github.com/theopenlab/spark/pull/19 We found the Spark QA
>>>>>> tests like
>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/ ,
>>>>>> it looks all tests seem never download the jar packages from maven centry
>>>>>> repo(such as
>>>>>> https://repo.maven.apache.org/maven2/org/opencypher/okapi-api/0.4.2/okapi-api-0.4.2.jar).
>>>>>> So we want to know how the jenkins jobs can do that, is there a internal
>>>>>> maven repo launched? maybe we can do the same thing to avoid the network
>>>>>> connection cost during downloading the dependent jar packages.
>>>>>>
>>>>>> Third, the most important thing, it's about ARM CI of spark, we
>>>>>> believe that it is necessary, right? And you can see we really made a lot
>>>>>> of efforts, now the basic arm build/test jobs is ok, so we suggest to add
>>>>>> arm jobs to community, we can set them to novoting firstly, and
>>>>>> improve/rich the jobs step by step. Generally, there are two ways in our
>>>>>> mind to integrate the ARM CI for spark:
>>>>>> 1) We introduce openlab ARM CI into spark as a custom CI system.
>>>>>> We provide human resources and test ARM VMs, also we will focus on the ARM
>>>>>> related issues about Spark. We will push the PR into community.
>>>>>> 2) We donate ARM VM resources into existing amplab Jenkins. We
>>>>>> still provide human resources, focus on the ARM related issues about Spark
>>>>>> and push the PR into community.
>>>>>> Both options, we will provide human resources to maintain, of course
>>>>>> it will be great if we can work together. So please tell us which option
>>>>>> you would like? And let's move forward. Waiting for your reply, thank you
>>>>>> very much.
>>>>>>
>>>>>
Re: Ask for ARM CI for spark
Posted by bo zhaobo <bz...@gmail.com>.
Hi Guys,
Recently, we are trying to test pyspark on ARM, we found some issue but
have no idea about them. Could you please have a look if you are free?
Thanks.
There are two issues:
1. The first one looks like a arm performance issue, the test job in a
pyspark test doesn't fully finish when exec assert check. So we change the
source code on our local env to test, they will pass. For this issue, we
opened a JIRA issue [1]. If you guys are free, please help it. Thanks.
2. The second one looks like a spark internal issue, when we test
"pyspark.mllib.tests.test_streaming_algorithms:StreamingLinearRegressionWithTests.test_train_prediction",
it will fail as the "condition" function.We tried to deep into it and found
the predicted value is still [0. 0. .....0.], eventhough we wait for a long
time on ARM testing env. That's the main cause I think. And we failed to
debug into which step is wrong. Could you please help to figure it out? I
upload the test log after I inserted some 'printf' into the 'func' function
of the testcase function. I tried on ARM and X86. ARM log is [2], X86 log
is [3]. They are the same testing env except different ARCH.
Thanks, if you are free, please help us.
Best Regards
[1] https://issues.apache.org/jira/browse/SPARK-29205
[2] https://etherpad.net/p/pyspark-arm
[3] https://etherpad.net/p/pyspark-x86
[image: Mailtrack]
<https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&>
Sender
notified by
Mailtrack
<https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&>
19/09/23
上午11:53:29
Tianhua huang <hu...@gmail.com> 于2019年9月19日周四 上午10:59写道：
> @Dongjoon Hyun <do...@gmail.com> ,
>
> Sure, and I have update the JIRA already :)
> https://issues.apache.org/jira/browse/SPARK-29106
> If anything missed, please let me know, thank you.
>
> On Thu, Sep 19, 2019 at 12:44 AM Dongjoon Hyun <do...@gmail.com>
> wrote:
>
>> Hi, Tianhua.
>>
>> Could you summarize the detail on the JIRA once more?
>> It will be very helpful for the community. Also, I've been waiting on
>> that JIRA. :)
>>
>> Bests,
>> Dongjoon.
>>
>>
>> On Mon, Sep 16, 2019 at 11:48 PM Tianhua huang <hu...@gmail.com>
>> wrote:
>>
>>> @shane knapp <sk...@berkeley.edu> thank you very much, I opened an
>>> issue for this https://issues.apache.org/jira/browse/SPARK-29106, we
>>> can tall the details in it :)
>>> And we will prepare an arm instance today and will send the info to your
>>> email later.
>>>
>>> On Tue, Sep 17, 2019 at 4:40 AM Shane Knapp <sk...@berkeley.edu> wrote:
>>>
>>>> @Tianhua huang <hu...@gmail.com> sure, i think we can get
>>>> something sorted for the short-term.
>>>>
>>>> all we need is ssh access (i can provide an ssh key), and i can then
>>>> have our jenkins master launch a remote worker on that instance.
>>>>
>>>> instance setup, etc, will be up to you. my support for the time being
>>>> will be to create the job and 'best effort' for everything else.
>>>>
>>>> this should get us up and running asap.
>>>>
>>>> is there an open JIRA for jenkins/arm test support? we can move the
>>>> technical details about this idea there.
>>>>
>>>> On Sun, Sep 15, 2019 at 9:03 PM Tianhua huang <
>>>> huangtianhua223@gmail.com> wrote:
>>>>
>>>>> @Sean Owen <sr...@gmail.com> , so sorry to reply late, we had a
>>>>> Mid-Autumn holiday:)
>>>>>
>>>>> If you hope to integrate ARM CI to amplab jenkins, we can offer the
>>>>> arm instance, and then the ARM job will run together with other x86 jobs,
>>>>> so maybe there is a guideline to do this? @shane knapp
>>>>> <sk...@berkeley.edu> would you help us?
>>>>>
>>>>> On Thu, Sep 12, 2019 at 9:36 PM Sean Owen <sr...@gmail.com> wrote:
>>>>>
>>>>>> I don't know what's involved in actually accepting or operating those
>>>>>> machines, so can't comment there, but in the meantime it's good that you
>>>>>> are running these tests and can help report changes needed to keep it
>>>>>> working with ARM. I would continue with that for now.
>>>>>>
>>>>>> On Wed, Sep 11, 2019 at 10:06 PM Tianhua huang <
>>>>>> huangtianhua223@gmail.com> wrote:
>>>>>>
>>>>>>> Hi all,
>>>>>>>
>>>>>>> For the whole work process of spark ARM CI, we want to make 2 things
>>>>>>> clear.
>>>>>>>
>>>>>>> The first thing is:
>>>>>>> About spark ARM CI, now we have two periodic jobs, one job[1] based
>>>>>>> on commit[2](which already fixed the replay tests failed issue[3], we made
>>>>>>> a new test branch based on date 09-09-2019), the other job[4] based on
>>>>>>> spark master.
>>>>>>>
>>>>>>> The first job we test on the specified branch to prove that our ARM
>>>>>>> CI is good and stable.
>>>>>>> The second job checks spark master every day, then we can find
>>>>>>> whether the latest commits affect the ARM CI. According to the build
>>>>>>> history and result, it shows that some problems are easier to find on ARM
>>>>>>> like SPARK-28770 <https://issues.apache.org/jira/browse/SPARK-28770>,
>>>>>>> and it also shows that we would make efforts to trace and figure them
>>>>>>> out, till now we have found and fixed several problems[5][6][7], thanks
>>>>>>> everyone of the community :). And we believe that ARM CI is very necessary,
>>>>>>> right?
>>>>>>>
>>>>>>> The second thing is:
>>>>>>> We plan to run the jobs for a period of time, and you can see the
>>>>>>> result and logs from 'build history' of the jobs console, if everything
>>>>>>> goes well for one or two weeks could community accept the ARM CI? or how
>>>>>>> long the periodic jobs to run then our community could have enough
>>>>>>> confidence to accept the ARM CI? As you suggested before, it's good to
>>>>>>> integrate ARM CI to amplab jenkins, we agree that and we can donate the ARM
>>>>>>> instances and then maintain the ARM-related test jobs together with
>>>>>>> community, any thoughts?
>>>>>>>
>>>>>>> Thank you all!
>>>>>>>
>>>>>>> [1]
>>>>>>> http://status.openlabtesting.org/job/spark-unchanged-branch-unit-test-hadoop-2.7-arm64
>>>>>>> [2]
>>>>>>> https://github.com/apache/spark/commit/0ed9fae45769d4b06b8cf8128f462f09ff3d9a72
>>>>>>> [3] https://issues.apache.org/jira/browse/SPARK-28770
>>>>>>> [4]
>>>>>>> http://status.openlabtesting.org/builds?job_name=spark-master-unit-test-hadoop-2.7-arm64
>>>>>>> [5] https://github.com/apache/spark/pull/25186
>>>>>>> [6] https://github.com/apache/spark/pull/25279
>>>>>>> [7] https://github.com/apache/spark/pull/25673
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> On Fri, Aug 16, 2019 at 11:24 PM Sean Owen <sr...@gmail.com> wrote:
>>>>>>>
>>>>>>>> Yes, I think it's just local caching. After you run the build you
>>>>>>>> should find lots of stuff cached at ~/.m2/repository and it won't download
>>>>>>>> every time.
>>>>>>>>
>>>>>>>> On Fri, Aug 16, 2019 at 3:01 AM bo zhaobo <
>>>>>>>> bzhaojyathousandy@gmail.com> wrote:
>>>>>>>>
>>>>>>>>> Hi Sean,
>>>>>>>>> Thanks for reply. And very apologize for making you confused.
>>>>>>>>> I know the dependencies will be downloaded from SBT or Maven. But
>>>>>>>>> the Spark QA job also exec "mvn clean package", why the log didn't print
>>>>>>>>> "downloading some jar from Maven central [1] and build very fast. Is the
>>>>>>>>> reason that Spark Jenkins build the Spark jars in the physical machiines
>>>>>>>>> and won't destrory the test env after job is finished? Then the other job
>>>>>>>>> build Spark will get the dependencies jar from the local cached, as the
>>>>>>>>> previous jobs exec "mvn package", those dependencies had been downloaded
>>>>>>>>> already on local worker machine. Am I right? Is that the reason the job
>>>>>>>>> log[1] didn't print any downloading information from Maven Central?
>>>>>>>>>
>>>>>>>>> Thank you very much.
>>>>>>>>>
>>>>>>>>> [1]
>>>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/job/spark-master-test-maven-hadoop-2.6-ubuntu-testing/lastBuild/consoleFull
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Best regards
>>>>>>>>>
>>>>>>>>> ZhaoBo
>>>>>>>>>
>>>>>>>>> [image: Mailtrack]
>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>>>>> notified by
>>>>>>>>> Mailtrack
>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/08/16
>>>>>>>>> 下午03:58:53
>>>>>>>>>
>>>>>>>>> Sean Owen <sr...@gmail.com> 于2019年8月16日周五 上午10:38写道：
>>>>>>>>>
>>>>>>>>>> I'm not sure what you mean. The dependencies are downloaded by
>>>>>>>>>> SBT and Maven like in any other project, and nothing about it is specific
>>>>>>>>>> to Spark.
>>>>>>>>>> The worker machines cache artifacts that are downloaded from
>>>>>>>>>> these, but this is a function of Maven and SBT, not Spark. You may find
>>>>>>>>>> that the initial download takes a long time.
>>>>>>>>>>
>>>>>>>>>> On Thu, Aug 15, 2019 at 9:02 PM bo zhaobo <
>>>>>>>>>> bzhaojyathousandy@gmail.com> wrote:
>>>>>>>>>>
>>>>>>>>>>> Hi Sean,
>>>>>>>>>>>
>>>>>>>>>>> Thanks very much for pointing out the roadmap. ;-). Then I think
>>>>>>>>>>> we will continue to focus on our test environment.
>>>>>>>>>>>
>>>>>>>>>>> For the networking problems, I mean that we can access Maven
>>>>>>>>>>> Central, and jobs cloud download the required jar package with a high
>>>>>>>>>>> network speed. What we want to know is that, why the Spark QA test jobs[1]
>>>>>>>>>>> log shows the job script/maven build seem don't download the jar packages?
>>>>>>>>>>> Could you tell us the reason about that? Thank you. The reason we raise
>>>>>>>>>>> the "networking problems" is that we found a phenomenon during we test, if
>>>>>>>>>>> we execute "mvn clean package" in a new test environment(As in our test
>>>>>>>>>>> environment, we will destory the test VMs after the job is finish), maven
>>>>>>>>>>> will download the dependency jar packages from Maven Central, but in this
>>>>>>>>>>> job "spark-master-test-maven-hadoop" [2], from the log, we didn't found it
>>>>>>>>>>> download any jar packages, what the reason about that?
>>>>>>>>>>> Also we build the Spark jar with downloading dependencies from
>>>>>>>>>>> Maven Central, it will cost mostly 1 hour. And we found [2] just cost
>>>>>>>>>>> 10min. But if we run "mvn package" in a VM which already exec "mvn package"
>>>>>>>>>>> before, it just cost 14min, looks very closer with [2]. So we suspect that
>>>>>>>>>>> downloading the Jar packages cost so much time. For the goad of ARM CI, we
>>>>>>>>>>> expect the performance of NEW ARM CI could be closer with existing X86 CI,
>>>>>>>>>>> then users could accept it eaiser.
>>>>>>>>>>>
>>>>>>>>>>> [1]
>>>>>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/
>>>>>>>>>>> [2]
>>>>>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/job/spark-master-test-maven-hadoop-2.6-ubuntu-testing/lastBuild/consoleFull
>>>>>>>>>>>
>>>>>>>>>>> Best regards
>>>>>>>>>>>
>>>>>>>>>>> ZhaoBo
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> [image: Mailtrack]
>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>>>>>>> notified by
>>>>>>>>>>> Mailtrack
>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/08/16
>>>>>>>>>>> 上午09:48:43
>>>>>>>>>>>
>>>>>>>>>>> Sean Owen <sr...@gmail.com> 于2019年8月15日周四 下午9:58写道：
>>>>>>>>>>>
>>>>>>>>>>>> I think the right goal is to fix the remaining issues first. If
>>>>>>>>>>>> we set up CI/CD it will only tell us there are still some test failures. If
>>>>>>>>>>>> it's stable, and not hard to add to the existing CI/CD, yes it could be
>>>>>>>>>>>> done automatically later. You can continue to test on ARM independently for
>>>>>>>>>>>> now.
>>>>>>>>>>>>
>>>>>>>>>>>> It sounds indeed like there are some networking problems in the
>>>>>>>>>>>> test system if you're not able to download from Maven Central. That rarely
>>>>>>>>>>>> takes significant time, and there aren't project-specific mirrors here. You
>>>>>>>>>>>> might be able to point at a closer public mirror, depending on where you
>>>>>>>>>>>> are.
>>>>>>>>>>>>
>>>>>>>>>>>> On Thu, Aug 15, 2019 at 5:43 AM Tianhua huang <
>>>>>>>>>>>> huangtianhua223@gmail.com> wrote:
>>>>>>>>>>>>
>>>>>>>>>>>>> Hi all,
>>>>>>>>>>>>>
>>>>>>>>>>>>> I want to discuss spark ARM CI again, we took some tests on
>>>>>>>>>>>>> arm instance based on master and the job includes
>>>>>>>>>>>>> https://github.com/theopenlab/spark/pull/13 and k8s
>>>>>>>>>>>>> integration https://github.com/theopenlab/spark/pull/17/ ,
>>>>>>>>>>>>> there are several things I want to talk about:
>>>>>>>>>>>>>
>>>>>>>>>>>>> First, about the failed tests:
>>>>>>>>>>>>> 1.we have fixed some problems like
>>>>>>>>>>>>> https://github.com/apache/spark/pull/25186 and
>>>>>>>>>>>>> https://github.com/apache/spark/pull/25279, thanks sean owen
>>>>>>>>>>>>> and others to help us.
>>>>>>>>>>>>> 2.we tried k8s integration test on arm, and met an error:
>>>>>>>>>>>>> apk fetch hangs, the tests passed after adding '--network host' option
>>>>>>>>>>>>> for command `docker build`, see:
>>>>>>>>>>>>>
>>>>>>>>>>>>> https://github.com/theopenlab/spark/pull/17/files#diff-5b731b14068240d63a93c393f6f9b1e8R176
>>>>>>>>>>>>> , the solution refers to
>>>>>>>>>>>>> https://github.com/gliderlabs/docker-alpine/issues/307 and I
>>>>>>>>>>>>> don't know whether it happened once in community CI, or maybe we should
>>>>>>>>>>>>> submit a pr to pass '--network host' when `docker build`?
>>>>>>>>>>>>> 3.we found there are two tests failed after the commit
>>>>>>>>>>>>> https://github.com/apache/spark/pull/23767 :
>>>>>>>>>>>>> ReplayListenerSuite:
>>>>>>>>>>>>> - ...
>>>>>>>>>>>>> - End-to-end replay *** FAILED ***
>>>>>>>>>>>>> "[driver]" did not equal "[1]"
>>>>>>>>>>>>> (JsonProtocolSuite.scala:622)
>>>>>>>>>>>>> - End-to-end replay with compression *** FAILED ***
>>>>>>>>>>>>> "[driver]" did not equal "[1]"
>>>>>>>>>>>>> (JsonProtocolSuite.scala:622)
>>>>>>>>>>>>>
>>>>>>>>>>>>> we tried to revert the commit and then the tests
>>>>>>>>>>>>> passed, the patch is too big and so sorry we can't find the reason till
>>>>>>>>>>>>> now, if you are interesting please try it, and it will be very appreciate
>>>>>>>>>>>>> if someone can help us to figure it out.
>>>>>>>>>>>>>
>>>>>>>>>>>>> Second, about the test time, we increased the flavor of arm
>>>>>>>>>>>>> instance to 16U16G, but seems there was no significant improvement, the k8s
>>>>>>>>>>>>> integration test took about one and a half hours, and the QA test(like
>>>>>>>>>>>>> spark-master-test-maven-hadoop-2.7 community jenkins job) took about
>>>>>>>>>>>>> seventeen hours(it is too long :(), we suspect that the reason is the
>>>>>>>>>>>>> performance and network,
>>>>>>>>>>>>> we split the jobs based on projects such as sql, core and so
>>>>>>>>>>>>> on, the time can be decrease to about seven hours, see
>>>>>>>>>>>>> https://github.com/theopenlab/spark/pull/19 We found the
>>>>>>>>>>>>> Spark QA tests like
>>>>>>>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/ ,
>>>>>>>>>>>>> it looks all tests seem never download the jar packages from maven centry
>>>>>>>>>>>>> repo(such as
>>>>>>>>>>>>> https://repo.maven.apache.org/maven2/org/opencypher/okapi-api/0.4.2/okapi-api-0.4.2.jar).
>>>>>>>>>>>>> So we want to know how the jenkins jobs can do that, is there a internal
>>>>>>>>>>>>> maven repo launched? maybe we can do the same thing to avoid the network
>>>>>>>>>>>>> connection cost during downloading the dependent jar packages.
>>>>>>>>>>>>>
>>>>>>>>>>>>> Third, the most important thing, it's about ARM CI of spark,
>>>>>>>>>>>>> we believe that it is necessary, right? And you can see we really made a
>>>>>>>>>>>>> lot of efforts, now the basic arm build/test jobs is ok, so we suggest to
>>>>>>>>>>>>> add arm jobs to community, we can set them to novoting firstly, and
>>>>>>>>>>>>> improve/rich the jobs step by step. Generally, there are two ways in our
>>>>>>>>>>>>> mind to integrate the ARM CI for spark:
>>>>>>>>>>>>> 1) We introduce openlab ARM CI into spark as a custom CI
>>>>>>>>>>>>> system. We provide human resources and test ARM VMs, also we will focus on
>>>>>>>>>>>>> the ARM related issues about Spark. We will push the PR into community.
>>>>>>>>>>>>> 2) We donate ARM VM resources into existing amplab
>>>>>>>>>>>>> Jenkins. We still provide human resources, focus on the ARM related issues
>>>>>>>>>>>>> about Spark and push the PR into community.
>>>>>>>>>>>>> Both options, we will provide human resources to maintain, of
>>>>>>>>>>>>> course it will be great if we can work together. So please tell us which
>>>>>>>>>>>>> option you would like? And let's move forward. Waiting for your reply,
>>>>>>>>>>>>> thank you very much.
>>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>
>>>> --
>>>> Shane Knapp
>>>> UC Berkeley EECS Research / RISELab Staff Technical Lead
>>>> https://rise.cs.berkeley.edu
>>>>
>>>
Re: Ask for ARM CI for spark
Posted by Tianhua huang <hu...@gmail.com>.
@Dongjoon Hyun <do...@gmail.com> ,
Sure, and I have update the JIRA already :)
https://issues.apache.org/jira/browse/SPARK-29106
If anything missed, please let me know, thank you.
On Thu, Sep 19, 2019 at 12:44 AM Dongjoon Hyun <do...@gmail.com>
wrote:
> Hi, Tianhua.
>
> Could you summarize the detail on the JIRA once more?
> It will be very helpful for the community. Also, I've been waiting on that
> JIRA. :)
>
> Bests,
> Dongjoon.
>
>
> On Mon, Sep 16, 2019 at 11:48 PM Tianhua huang <hu...@gmail.com>
> wrote:
>
>> @shane knapp <sk...@berkeley.edu> thank you very much, I opened an
>> issue for this https://issues.apache.org/jira/browse/SPARK-29106, we can
>> tall the details in it :)
>> And we will prepare an arm instance today and will send the info to your
>> email later.
>>
>> On Tue, Sep 17, 2019 at 4:40 AM Shane Knapp <sk...@berkeley.edu> wrote:
>>
>>> @Tianhua huang <hu...@gmail.com> sure, i think we can get
>>> something sorted for the short-term.
>>>
>>> all we need is ssh access (i can provide an ssh key), and i can then
>>> have our jenkins master launch a remote worker on that instance.
>>>
>>> instance setup, etc, will be up to you. my support for the time being
>>> will be to create the job and 'best effort' for everything else.
>>>
>>> this should get us up and running asap.
>>>
>>> is there an open JIRA for jenkins/arm test support? we can move the
>>> technical details about this idea there.
>>>
>>> On Sun, Sep 15, 2019 at 9:03 PM Tianhua huang <hu...@gmail.com>
>>> wrote:
>>>
>>>> @Sean Owen <sr...@gmail.com> , so sorry to reply late, we had a
>>>> Mid-Autumn holiday:)
>>>>
>>>> If you hope to integrate ARM CI to amplab jenkins, we can offer the arm
>>>> instance, and then the ARM job will run together with other x86 jobs, so
>>>> maybe there is a guideline to do this? @shane knapp
>>>> <sk...@berkeley.edu> would you help us?
>>>>
>>>> On Thu, Sep 12, 2019 at 9:36 PM Sean Owen <sr...@gmail.com> wrote:
>>>>
>>>>> I don't know what's involved in actually accepting or operating those
>>>>> machines, so can't comment there, but in the meantime it's good that you
>>>>> are running these tests and can help report changes needed to keep it
>>>>> working with ARM. I would continue with that for now.
>>>>>
>>>>> On Wed, Sep 11, 2019 at 10:06 PM Tianhua huang <
>>>>> huangtianhua223@gmail.com> wrote:
>>>>>
>>>>>> Hi all,
>>>>>>
>>>>>> For the whole work process of spark ARM CI, we want to make 2 things
>>>>>> clear.
>>>>>>
>>>>>> The first thing is:
>>>>>> About spark ARM CI, now we have two periodic jobs, one job[1] based
>>>>>> on commit[2](which already fixed the replay tests failed issue[3], we made
>>>>>> a new test branch based on date 09-09-2019), the other job[4] based on
>>>>>> spark master.
>>>>>>
>>>>>> The first job we test on the specified branch to prove that our ARM
>>>>>> CI is good and stable.
>>>>>> The second job checks spark master every day, then we can find
>>>>>> whether the latest commits affect the ARM CI. According to the build
>>>>>> history and result, it shows that some problems are easier to find on ARM
>>>>>> like SPARK-28770 <https://issues.apache.org/jira/browse/SPARK-28770>,
>>>>>> and it also shows that we would make efforts to trace and figure them
>>>>>> out, till now we have found and fixed several problems[5][6][7], thanks
>>>>>> everyone of the community :). And we believe that ARM CI is very necessary,
>>>>>> right?
>>>>>>
>>>>>> The second thing is:
>>>>>> We plan to run the jobs for a period of time, and you can see the
>>>>>> result and logs from 'build history' of the jobs console, if everything
>>>>>> goes well for one or two weeks could community accept the ARM CI? or how
>>>>>> long the periodic jobs to run then our community could have enough
>>>>>> confidence to accept the ARM CI? As you suggested before, it's good to
>>>>>> integrate ARM CI to amplab jenkins, we agree that and we can donate the ARM
>>>>>> instances and then maintain the ARM-related test jobs together with
>>>>>> community, any thoughts?
>>>>>>
>>>>>> Thank you all!
>>>>>>
>>>>>> [1]
>>>>>> http://status.openlabtesting.org/job/spark-unchanged-branch-unit-test-hadoop-2.7-arm64
>>>>>> [2]
>>>>>> https://github.com/apache/spark/commit/0ed9fae45769d4b06b8cf8128f462f09ff3d9a72
>>>>>> [3] https://issues.apache.org/jira/browse/SPARK-28770
>>>>>> [4]
>>>>>> http://status.openlabtesting.org/builds?job_name=spark-master-unit-test-hadoop-2.7-arm64
>>>>>> [5] https://github.com/apache/spark/pull/25186
>>>>>> [6] https://github.com/apache/spark/pull/25279
>>>>>> [7] https://github.com/apache/spark/pull/25673
>>>>>>
>>>>>>
>>>>>>
>>>>>> On Fri, Aug 16, 2019 at 11:24 PM Sean Owen <sr...@gmail.com> wrote:
>>>>>>
>>>>>>> Yes, I think it's just local caching. After you run the build you
>>>>>>> should find lots of stuff cached at ~/.m2/repository and it won't download
>>>>>>> every time.
>>>>>>>
>>>>>>> On Fri, Aug 16, 2019 at 3:01 AM bo zhaobo <
>>>>>>> bzhaojyathousandy@gmail.com> wrote:
>>>>>>>
>>>>>>>> Hi Sean,
>>>>>>>> Thanks for reply. And very apologize for making you confused.
>>>>>>>> I know the dependencies will be downloaded from SBT or Maven. But
>>>>>>>> the Spark QA job also exec "mvn clean package", why the log didn't print
>>>>>>>> "downloading some jar from Maven central [1] and build very fast. Is the
>>>>>>>> reason that Spark Jenkins build the Spark jars in the physical machiines
>>>>>>>> and won't destrory the test env after job is finished? Then the other job
>>>>>>>> build Spark will get the dependencies jar from the local cached, as the
>>>>>>>> previous jobs exec "mvn package", those dependencies had been downloaded
>>>>>>>> already on local worker machine. Am I right? Is that the reason the job
>>>>>>>> log[1] didn't print any downloading information from Maven Central?
>>>>>>>>
>>>>>>>> Thank you very much.
>>>>>>>>
>>>>>>>> [1]
>>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/job/spark-master-test-maven-hadoop-2.6-ubuntu-testing/lastBuild/consoleFull
>>>>>>>>
>>>>>>>>
>>>>>>>> Best regards
>>>>>>>>
>>>>>>>> ZhaoBo
>>>>>>>>
>>>>>>>> [image: Mailtrack]
>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>>>> notified by
>>>>>>>> Mailtrack
>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/08/16
>>>>>>>> 下午03:58:53
>>>>>>>>
>>>>>>>> Sean Owen <sr...@gmail.com> 于2019年8月16日周五 上午10:38写道：
>>>>>>>>
>>>>>>>>> I'm not sure what you mean. The dependencies are downloaded by SBT
>>>>>>>>> and Maven like in any other project, and nothing about it is specific to
>>>>>>>>> Spark.
>>>>>>>>> The worker machines cache artifacts that are downloaded from
>>>>>>>>> these, but this is a function of Maven and SBT, not Spark. You may find
>>>>>>>>> that the initial download takes a long time.
>>>>>>>>>
>>>>>>>>> On Thu, Aug 15, 2019 at 9:02 PM bo zhaobo <
>>>>>>>>> bzhaojyathousandy@gmail.com> wrote:
>>>>>>>>>
>>>>>>>>>> Hi Sean,
>>>>>>>>>>
>>>>>>>>>> Thanks very much for pointing out the roadmap. ;-). Then I think
>>>>>>>>>> we will continue to focus on our test environment.
>>>>>>>>>>
>>>>>>>>>> For the networking problems, I mean that we can access Maven
>>>>>>>>>> Central, and jobs cloud download the required jar package with a high
>>>>>>>>>> network speed. What we want to know is that, why the Spark QA test jobs[1]
>>>>>>>>>> log shows the job script/maven build seem don't download the jar packages?
>>>>>>>>>> Could you tell us the reason about that? Thank you. The reason we raise
>>>>>>>>>> the "networking problems" is that we found a phenomenon during we test, if
>>>>>>>>>> we execute "mvn clean package" in a new test environment(As in our test
>>>>>>>>>> environment, we will destory the test VMs after the job is finish), maven
>>>>>>>>>> will download the dependency jar packages from Maven Central, but in this
>>>>>>>>>> job "spark-master-test-maven-hadoop" [2], from the log, we didn't found it
>>>>>>>>>> download any jar packages, what the reason about that?
>>>>>>>>>> Also we build the Spark jar with downloading dependencies from
>>>>>>>>>> Maven Central, it will cost mostly 1 hour. And we found [2] just cost
>>>>>>>>>> 10min. But if we run "mvn package" in a VM which already exec "mvn package"
>>>>>>>>>> before, it just cost 14min, looks very closer with [2]. So we suspect that
>>>>>>>>>> downloading the Jar packages cost so much time. For the goad of ARM CI, we
>>>>>>>>>> expect the performance of NEW ARM CI could be closer with existing X86 CI,
>>>>>>>>>> then users could accept it eaiser.
>>>>>>>>>>
>>>>>>>>>> [1]
>>>>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/
>>>>>>>>>> [2]
>>>>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/job/spark-master-test-maven-hadoop-2.6-ubuntu-testing/lastBuild/consoleFull
>>>>>>>>>>
>>>>>>>>>> Best regards
>>>>>>>>>>
>>>>>>>>>> ZhaoBo
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> [image: Mailtrack]
>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>>>>>> notified by
>>>>>>>>>> Mailtrack
>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/08/16
>>>>>>>>>> 上午09:48:43
>>>>>>>>>>
>>>>>>>>>> Sean Owen <sr...@gmail.com> 于2019年8月15日周四 下午9:58写道：
>>>>>>>>>>
>>>>>>>>>>> I think the right goal is to fix the remaining issues first. If
>>>>>>>>>>> we set up CI/CD it will only tell us there are still some test failures. If
>>>>>>>>>>> it's stable, and not hard to add to the existing CI/CD, yes it could be
>>>>>>>>>>> done automatically later. You can continue to test on ARM independently for
>>>>>>>>>>> now.
>>>>>>>>>>>
>>>>>>>>>>> It sounds indeed like there are some networking problems in the
>>>>>>>>>>> test system if you're not able to download from Maven Central. That rarely
>>>>>>>>>>> takes significant time, and there aren't project-specific mirrors here. You
>>>>>>>>>>> might be able to point at a closer public mirror, depending on where you
>>>>>>>>>>> are.
>>>>>>>>>>>
>>>>>>>>>>> On Thu, Aug 15, 2019 at 5:43 AM Tianhua huang <
>>>>>>>>>>> huangtianhua223@gmail.com> wrote:
>>>>>>>>>>>
>>>>>>>>>>>> Hi all,
>>>>>>>>>>>>
>>>>>>>>>>>> I want to discuss spark ARM CI again, we took some tests on arm
>>>>>>>>>>>> instance based on master and the job includes
>>>>>>>>>>>> https://github.com/theopenlab/spark/pull/13 and k8s
>>>>>>>>>>>> integration https://github.com/theopenlab/spark/pull/17/ ,
>>>>>>>>>>>> there are several things I want to talk about:
>>>>>>>>>>>>
>>>>>>>>>>>> First, about the failed tests:
>>>>>>>>>>>> 1.we have fixed some problems like
>>>>>>>>>>>> https://github.com/apache/spark/pull/25186 and
>>>>>>>>>>>> https://github.com/apache/spark/pull/25279, thanks sean owen
>>>>>>>>>>>> and others to help us.
>>>>>>>>>>>> 2.we tried k8s integration test on arm, and met an error:
>>>>>>>>>>>> apk fetch hangs, the tests passed after adding '--network host' option
>>>>>>>>>>>> for command `docker build`, see:
>>>>>>>>>>>>
>>>>>>>>>>>> https://github.com/theopenlab/spark/pull/17/files#diff-5b731b14068240d63a93c393f6f9b1e8R176
>>>>>>>>>>>> , the solution refers to
>>>>>>>>>>>> https://github.com/gliderlabs/docker-alpine/issues/307 and I
>>>>>>>>>>>> don't know whether it happened once in community CI, or maybe we should
>>>>>>>>>>>> submit a pr to pass '--network host' when `docker build`?
>>>>>>>>>>>> 3.we found there are two tests failed after the commit
>>>>>>>>>>>> https://github.com/apache/spark/pull/23767 :
>>>>>>>>>>>> ReplayListenerSuite:
>>>>>>>>>>>> - ...
>>>>>>>>>>>> - End-to-end replay *** FAILED ***
>>>>>>>>>>>> "[driver]" did not equal "[1]"
>>>>>>>>>>>> (JsonProtocolSuite.scala:622)
>>>>>>>>>>>> - End-to-end replay with compression *** FAILED ***
>>>>>>>>>>>> "[driver]" did not equal "[1]"
>>>>>>>>>>>> (JsonProtocolSuite.scala:622)
>>>>>>>>>>>>
>>>>>>>>>>>> we tried to revert the commit and then the tests
>>>>>>>>>>>> passed, the patch is too big and so sorry we can't find the reason till
>>>>>>>>>>>> now, if you are interesting please try it, and it will be very appreciate
>>>>>>>>>>>> if someone can help us to figure it out.
>>>>>>>>>>>>
>>>>>>>>>>>> Second, about the test time, we increased the flavor of arm
>>>>>>>>>>>> instance to 16U16G, but seems there was no significant improvement, the k8s
>>>>>>>>>>>> integration test took about one and a half hours, and the QA test(like
>>>>>>>>>>>> spark-master-test-maven-hadoop-2.7 community jenkins job) took about
>>>>>>>>>>>> seventeen hours(it is too long :(), we suspect that the reason is the
>>>>>>>>>>>> performance and network,
>>>>>>>>>>>> we split the jobs based on projects such as sql, core and so
>>>>>>>>>>>> on, the time can be decrease to about seven hours, see
>>>>>>>>>>>> https://github.com/theopenlab/spark/pull/19 We found the Spark
>>>>>>>>>>>> QA tests like
>>>>>>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/ ,
>>>>>>>>>>>> it looks all tests seem never download the jar packages from maven centry
>>>>>>>>>>>> repo(such as
>>>>>>>>>>>> https://repo.maven.apache.org/maven2/org/opencypher/okapi-api/0.4.2/okapi-api-0.4.2.jar).
>>>>>>>>>>>> So we want to know how the jenkins jobs can do that, is there a internal
>>>>>>>>>>>> maven repo launched? maybe we can do the same thing to avoid the network
>>>>>>>>>>>> connection cost during downloading the dependent jar packages.
>>>>>>>>>>>>
>>>>>>>>>>>> Third, the most important thing, it's about ARM CI of spark, we
>>>>>>>>>>>> believe that it is necessary, right? And you can see we really made a lot
>>>>>>>>>>>> of efforts, now the basic arm build/test jobs is ok, so we suggest to add
>>>>>>>>>>>> arm jobs to community, we can set them to novoting firstly, and
>>>>>>>>>>>> improve/rich the jobs step by step. Generally, there are two ways in our
>>>>>>>>>>>> mind to integrate the ARM CI for spark:
>>>>>>>>>>>> 1) We introduce openlab ARM CI into spark as a custom CI
>>>>>>>>>>>> system. We provide human resources and test ARM VMs, also we will focus on
>>>>>>>>>>>> the ARM related issues about Spark. We will push the PR into community.
>>>>>>>>>>>> 2) We donate ARM VM resources into existing amplab
>>>>>>>>>>>> Jenkins. We still provide human resources, focus on the ARM related issues
>>>>>>>>>>>> about Spark and push the PR into community.
>>>>>>>>>>>> Both options, we will provide human resources to maintain, of
>>>>>>>>>>>> course it will be great if we can work together. So please tell us which
>>>>>>>>>>>> option you would like? And let's move forward. Waiting for your reply,
>>>>>>>>>>>> thank you very much.
>>>>>>>>>>>>
>>>>>>>>>>>
>>>
>>> --
>>> Shane Knapp
>>> UC Berkeley EECS Research / RISELab Staff Technical Lead
>>> https://rise.cs.berkeley.edu
>>>
>>
Re: Ask for ARM CI for spark
Posted by Dongjoon Hyun <do...@gmail.com>.
Hi, Tianhua.
Could you summarize the detail on the JIRA once more?
It will be very helpful for the community. Also, I've been waiting on that
JIRA. :)
Bests,
Dongjoon.
On Mon, Sep 16, 2019 at 11:48 PM Tianhua huang <hu...@gmail.com>
wrote:
> @shane knapp <sk...@berkeley.edu> thank you very much, I opened an issue
> for this https://issues.apache.org/jira/browse/SPARK-29106, we can tall
> the details in it :)
> And we will prepare an arm instance today and will send the info to your
> email later.
>
> On Tue, Sep 17, 2019 at 4:40 AM Shane Knapp <sk...@berkeley.edu> wrote:
>
>> @Tianhua huang <hu...@gmail.com> sure, i think we can get
>> something sorted for the short-term.
>>
>> all we need is ssh access (i can provide an ssh key), and i can then have
>> our jenkins master launch a remote worker on that instance.
>>
>> instance setup, etc, will be up to you. my support for the time being
>> will be to create the job and 'best effort' for everything else.
>>
>> this should get us up and running asap.
>>
>> is there an open JIRA for jenkins/arm test support? we can move the
>> technical details about this idea there.
>>
>> On Sun, Sep 15, 2019 at 9:03 PM Tianhua huang <hu...@gmail.com>
>> wrote:
>>
>>> @Sean Owen <sr...@gmail.com> , so sorry to reply late, we had a
>>> Mid-Autumn holiday:)
>>>
>>> If you hope to integrate ARM CI to amplab jenkins, we can offer the arm
>>> instance, and then the ARM job will run together with other x86 jobs, so
>>> maybe there is a guideline to do this? @shane knapp
>>> <sk...@berkeley.edu> would you help us?
>>>
>>> On Thu, Sep 12, 2019 at 9:36 PM Sean Owen <sr...@gmail.com> wrote:
>>>
>>>> I don't know what's involved in actually accepting or operating those
>>>> machines, so can't comment there, but in the meantime it's good that you
>>>> are running these tests and can help report changes needed to keep it
>>>> working with ARM. I would continue with that for now.
>>>>
>>>> On Wed, Sep 11, 2019 at 10:06 PM Tianhua huang <
>>>> huangtianhua223@gmail.com> wrote:
>>>>
>>>>> Hi all,
>>>>>
>>>>> For the whole work process of spark ARM CI, we want to make 2 things
>>>>> clear.
>>>>>
>>>>> The first thing is:
>>>>> About spark ARM CI, now we have two periodic jobs, one job[1] based on
>>>>> commit[2](which already fixed the replay tests failed issue[3], we made a
>>>>> new test branch based on date 09-09-2019), the other job[4] based on spark
>>>>> master.
>>>>>
>>>>> The first job we test on the specified branch to prove that our ARM CI
>>>>> is good and stable.
>>>>> The second job checks spark master every day, then we can find whether
>>>>> the latest commits affect the ARM CI. According to the build history and
>>>>> result, it shows that some problems are easier to find on ARM like
>>>>> SPARK-28770 <https://issues.apache.org/jira/browse/SPARK-28770>, and
>>>>> it also shows that we would make efforts to trace and figure them out, till
>>>>> now we have found and fixed several problems[5][6][7], thanks everyone of
>>>>> the community :). And we believe that ARM CI is very necessary, right?
>>>>>
>>>>> The second thing is:
>>>>> We plan to run the jobs for a period of time, and you can see the
>>>>> result and logs from 'build history' of the jobs console, if everything
>>>>> goes well for one or two weeks could community accept the ARM CI? or how
>>>>> long the periodic jobs to run then our community could have enough
>>>>> confidence to accept the ARM CI? As you suggested before, it's good to
>>>>> integrate ARM CI to amplab jenkins, we agree that and we can donate the ARM
>>>>> instances and then maintain the ARM-related test jobs together with
>>>>> community, any thoughts?
>>>>>
>>>>> Thank you all!
>>>>>
>>>>> [1]
>>>>> http://status.openlabtesting.org/job/spark-unchanged-branch-unit-test-hadoop-2.7-arm64
>>>>> [2]
>>>>> https://github.com/apache/spark/commit/0ed9fae45769d4b06b8cf8128f462f09ff3d9a72
>>>>> [3] https://issues.apache.org/jira/browse/SPARK-28770
>>>>> [4]
>>>>> http://status.openlabtesting.org/builds?job_name=spark-master-unit-test-hadoop-2.7-arm64
>>>>> [5] https://github.com/apache/spark/pull/25186
>>>>> [6] https://github.com/apache/spark/pull/25279
>>>>> [7] https://github.com/apache/spark/pull/25673
>>>>>
>>>>>
>>>>>
>>>>> On Fri, Aug 16, 2019 at 11:24 PM Sean Owen <sr...@gmail.com> wrote:
>>>>>
>>>>>> Yes, I think it's just local caching. After you run the build you
>>>>>> should find lots of stuff cached at ~/.m2/repository and it won't download
>>>>>> every time.
>>>>>>
>>>>>> On Fri, Aug 16, 2019 at 3:01 AM bo zhaobo <
>>>>>> bzhaojyathousandy@gmail.com> wrote:
>>>>>>
>>>>>>> Hi Sean,
>>>>>>> Thanks for reply. And very apologize for making you confused.
>>>>>>> I know the dependencies will be downloaded from SBT or Maven. But
>>>>>>> the Spark QA job also exec "mvn clean package", why the log didn't print
>>>>>>> "downloading some jar from Maven central [1] and build very fast. Is the
>>>>>>> reason that Spark Jenkins build the Spark jars in the physical machiines
>>>>>>> and won't destrory the test env after job is finished? Then the other job
>>>>>>> build Spark will get the dependencies jar from the local cached, as the
>>>>>>> previous jobs exec "mvn package", those dependencies had been downloaded
>>>>>>> already on local worker machine. Am I right? Is that the reason the job
>>>>>>> log[1] didn't print any downloading information from Maven Central?
>>>>>>>
>>>>>>> Thank you very much.
>>>>>>>
>>>>>>> [1]
>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/job/spark-master-test-maven-hadoop-2.6-ubuntu-testing/lastBuild/consoleFull
>>>>>>>
>>>>>>>
>>>>>>> Best regards
>>>>>>>
>>>>>>> ZhaoBo
>>>>>>>
>>>>>>> [image: Mailtrack]
>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>>> notified by
>>>>>>> Mailtrack
>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/08/16
>>>>>>> 下午03:58:53
>>>>>>>
>>>>>>> Sean Owen <sr...@gmail.com> 于2019年8月16日周五 上午10:38写道：
>>>>>>>
>>>>>>>> I'm not sure what you mean. The dependencies are downloaded by SBT
>>>>>>>> and Maven like in any other project, and nothing about it is specific to
>>>>>>>> Spark.
>>>>>>>> The worker machines cache artifacts that are downloaded from these,
>>>>>>>> but this is a function of Maven and SBT, not Spark. You may find that the
>>>>>>>> initial download takes a long time.
>>>>>>>>
>>>>>>>> On Thu, Aug 15, 2019 at 9:02 PM bo zhaobo <
>>>>>>>> bzhaojyathousandy@gmail.com> wrote:
>>>>>>>>
>>>>>>>>> Hi Sean,
>>>>>>>>>
>>>>>>>>> Thanks very much for pointing out the roadmap. ;-). Then I think
>>>>>>>>> we will continue to focus on our test environment.
>>>>>>>>>
>>>>>>>>> For the networking problems, I mean that we can access Maven
>>>>>>>>> Central, and jobs cloud download the required jar package with a high
>>>>>>>>> network speed. What we want to know is that, why the Spark QA test jobs[1]
>>>>>>>>> log shows the job script/maven build seem don't download the jar packages?
>>>>>>>>> Could you tell us the reason about that? Thank you. The reason we raise
>>>>>>>>> the "networking problems" is that we found a phenomenon during we test, if
>>>>>>>>> we execute "mvn clean package" in a new test environment(As in our test
>>>>>>>>> environment, we will destory the test VMs after the job is finish), maven
>>>>>>>>> will download the dependency jar packages from Maven Central, but in this
>>>>>>>>> job "spark-master-test-maven-hadoop" [2], from the log, we didn't found it
>>>>>>>>> download any jar packages, what the reason about that?
>>>>>>>>> Also we build the Spark jar with downloading dependencies from
>>>>>>>>> Maven Central, it will cost mostly 1 hour. And we found [2] just cost
>>>>>>>>> 10min. But if we run "mvn package" in a VM which already exec "mvn package"
>>>>>>>>> before, it just cost 14min, looks very closer with [2]. So we suspect that
>>>>>>>>> downloading the Jar packages cost so much time. For the goad of ARM CI, we
>>>>>>>>> expect the performance of NEW ARM CI could be closer with existing X86 CI,
>>>>>>>>> then users could accept it eaiser.
>>>>>>>>>
>>>>>>>>> [1] https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/
>>>>>>>>>
>>>>>>>>> [2]
>>>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/job/spark-master-test-maven-hadoop-2.6-ubuntu-testing/lastBuild/consoleFull
>>>>>>>>>
>>>>>>>>> Best regards
>>>>>>>>>
>>>>>>>>> ZhaoBo
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> [image: Mailtrack]
>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>>>>> notified by
>>>>>>>>> Mailtrack
>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/08/16
>>>>>>>>> 上午09:48:43
>>>>>>>>>
>>>>>>>>> Sean Owen <sr...@gmail.com> 于2019年8月15日周四 下午9:58写道：
>>>>>>>>>
>>>>>>>>>> I think the right goal is to fix the remaining issues first. If
>>>>>>>>>> we set up CI/CD it will only tell us there are still some test failures. If
>>>>>>>>>> it's stable, and not hard to add to the existing CI/CD, yes it could be
>>>>>>>>>> done automatically later. You can continue to test on ARM independently for
>>>>>>>>>> now.
>>>>>>>>>>
>>>>>>>>>> It sounds indeed like there are some networking problems in the
>>>>>>>>>> test system if you're not able to download from Maven Central. That rarely
>>>>>>>>>> takes significant time, and there aren't project-specific mirrors here. You
>>>>>>>>>> might be able to point at a closer public mirror, depending on where you
>>>>>>>>>> are.
>>>>>>>>>>
>>>>>>>>>> On Thu, Aug 15, 2019 at 5:43 AM Tianhua huang <
>>>>>>>>>> huangtianhua223@gmail.com> wrote:
>>>>>>>>>>
>>>>>>>>>>> Hi all,
>>>>>>>>>>>
>>>>>>>>>>> I want to discuss spark ARM CI again, we took some tests on arm
>>>>>>>>>>> instance based on master and the job includes
>>>>>>>>>>> https://github.com/theopenlab/spark/pull/13 and k8s
>>>>>>>>>>> integration https://github.com/theopenlab/spark/pull/17/ ,
>>>>>>>>>>> there are several things I want to talk about:
>>>>>>>>>>>
>>>>>>>>>>> First, about the failed tests:
>>>>>>>>>>> 1.we have fixed some problems like
>>>>>>>>>>> https://github.com/apache/spark/pull/25186 and
>>>>>>>>>>> https://github.com/apache/spark/pull/25279, thanks sean owen
>>>>>>>>>>> and others to help us.
>>>>>>>>>>> 2.we tried k8s integration test on arm, and met an error:
>>>>>>>>>>> apk fetch hangs, the tests passed after adding '--network host' option
>>>>>>>>>>> for command `docker build`, see:
>>>>>>>>>>>
>>>>>>>>>>> https://github.com/theopenlab/spark/pull/17/files#diff-5b731b14068240d63a93c393f6f9b1e8R176
>>>>>>>>>>> , the solution refers to
>>>>>>>>>>> https://github.com/gliderlabs/docker-alpine/issues/307 and I
>>>>>>>>>>> don't know whether it happened once in community CI, or maybe we should
>>>>>>>>>>> submit a pr to pass '--network host' when `docker build`?
>>>>>>>>>>> 3.we found there are two tests failed after the commit
>>>>>>>>>>> https://github.com/apache/spark/pull/23767 :
>>>>>>>>>>> ReplayListenerSuite:
>>>>>>>>>>> - ...
>>>>>>>>>>> - End-to-end replay *** FAILED ***
>>>>>>>>>>> "[driver]" did not equal "[1]"
>>>>>>>>>>> (JsonProtocolSuite.scala:622)
>>>>>>>>>>> - End-to-end replay with compression *** FAILED ***
>>>>>>>>>>> "[driver]" did not equal "[1]"
>>>>>>>>>>> (JsonProtocolSuite.scala:622)
>>>>>>>>>>>
>>>>>>>>>>> we tried to revert the commit and then the tests passed,
>>>>>>>>>>> the patch is too big and so sorry we can't find the reason till now, if you
>>>>>>>>>>> are interesting please try it, and it will be very appreciate if
>>>>>>>>>>> someone can help us to figure it out.
>>>>>>>>>>>
>>>>>>>>>>> Second, about the test time, we increased the flavor of arm
>>>>>>>>>>> instance to 16U16G, but seems there was no significant improvement, the k8s
>>>>>>>>>>> integration test took about one and a half hours, and the QA test(like
>>>>>>>>>>> spark-master-test-maven-hadoop-2.7 community jenkins job) took about
>>>>>>>>>>> seventeen hours(it is too long :(), we suspect that the reason is the
>>>>>>>>>>> performance and network,
>>>>>>>>>>> we split the jobs based on projects such as sql, core and so on,
>>>>>>>>>>> the time can be decrease to about seven hours, see
>>>>>>>>>>> https://github.com/theopenlab/spark/pull/19 We found the Spark
>>>>>>>>>>> QA tests like
>>>>>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/ ,
>>>>>>>>>>> it looks all tests seem never download the jar packages from maven centry
>>>>>>>>>>> repo(such as
>>>>>>>>>>> https://repo.maven.apache.org/maven2/org/opencypher/okapi-api/0.4.2/okapi-api-0.4.2.jar).
>>>>>>>>>>> So we want to know how the jenkins jobs can do that, is there a internal
>>>>>>>>>>> maven repo launched? maybe we can do the same thing to avoid the network
>>>>>>>>>>> connection cost during downloading the dependent jar packages.
>>>>>>>>>>>
>>>>>>>>>>> Third, the most important thing, it's about ARM CI of spark, we
>>>>>>>>>>> believe that it is necessary, right? And you can see we really made a lot
>>>>>>>>>>> of efforts, now the basic arm build/test jobs is ok, so we suggest to add
>>>>>>>>>>> arm jobs to community, we can set them to novoting firstly, and
>>>>>>>>>>> improve/rich the jobs step by step. Generally, there are two ways in our
>>>>>>>>>>> mind to integrate the ARM CI for spark:
>>>>>>>>>>> 1) We introduce openlab ARM CI into spark as a custom CI
>>>>>>>>>>> system. We provide human resources and test ARM VMs, also we will focus on
>>>>>>>>>>> the ARM related issues about Spark. We will push the PR into community.
>>>>>>>>>>> 2) We donate ARM VM resources into existing amplab Jenkins.
>>>>>>>>>>> We still provide human resources, focus on the ARM related issues about
>>>>>>>>>>> Spark and push the PR into community.
>>>>>>>>>>> Both options, we will provide human resources to maintain, of
>>>>>>>>>>> course it will be great if we can work together. So please tell us which
>>>>>>>>>>> option you would like? And let's move forward. Waiting for your reply,
>>>>>>>>>>> thank you very much.
>>>>>>>>>>>
>>>>>>>>>>
>>
>> --
>> Shane Knapp
>> UC Berkeley EECS Research / RISELab Staff Technical Lead
>> https://rise.cs.berkeley.edu
>>
>
Re: Ask for ARM CI for spark
Posted by Tianhua huang <hu...@gmail.com>.
We can talk about this later, but I have to update some things:)
- It (largely) worked previously
--- But no one sure about this before the arm testing, and it can't be
found anywhere, specify officially will make it more clear
- I think you're also saying you don't have 100% tests passing anyway,
though probably just small issues
--- The maven and python tests are 100% passing, see
https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-maven-arm/ and
https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-python-arm/
- It does not seem to merit a special announcement from the PMC among
the 2000+ changes in Spark 3
--- It's import to users, I believe it deserves
On Mon, Nov 18, 2019 at 10:06 AM Sean Owen <sr...@apache.org> wrote:
> Same response as before:
>
> - It is in the list of resolved JIRAs, of course
> - It (largely) worked previously
> - I think you're also saying you don't have 100% tests passing anyway,
> though probably just small issues
> - It does not seem to merit a special announcement from the PMC among
> the 2000+ changes in Spark 3
> - You are welcome to announce (on the project's user@ list if you
> like) whatever you want. Obviously, this is already well advertised on
> dev@
>
> I think you are asking for what borders on endorsement, and no that
> doesn't sound appropriate. Please just announce whatever you like as
> suggested.
>
> Sean
>
> On Sun, Nov 17, 2019 at 8:01 PM Tianhua huang <hu...@gmail.com>
> wrote:
> >
> > @Sean Owen，
> > I'm afraid I don't agree with you this time, I still remember no one can
> tell me whether Spark supports ARM or how much Spark can support ARM when I
> asked this first time on Dev@, you're very kind and told me to build and
> test on ARM locally and so sorry I think you were not sure much about this
> at that moment, right? Then I and my team work with community, we
> found/fixed several issues, integrate arm jobs into AMPLAB Jenkins, and the
> daily jobs has been stablely running for few weeks... after these efforts
> why not announce this officially in Spark releasenote? I believe after this
> everyone will know Spark is fully testing on ARM on community CI, Spark
> supports ARM basically, it's amazing and this will be very helpful. So what
> do you think? Or what are you worrying about?
>
Re: Ask for ARM CI for spark
Posted by bo zhaobo <bz...@gmail.com>.
Hi @Sean Owen <sr...@gmail.com> ,
Thanks for your reply and patient.
First, we are so apologized for the bad words in the previous emails. We
just want to make the users can see the current support status in some
place of spark community. I'm really appreciated that you and spark
community make spark better on ARM, and open to us.
Second, that's correct, adding this kind information of CICD into
releasenotes is improper. So we won't do that.
Third, I think your suggest is good for current situation. We will follow
your kind suggestion and send a email to user@ and dev@ to describe our
testing result, including the test coverage and the known issue we found.
Also we hope that we do this could be good for attracting more
users/developers of spark use ARM.
Fourth, we still have concerns that user still can not know very clear that
Spark can run on which ARCH with generic testing. So user will ask the same
question again and again which mentioned by huangtianhua in spark
community, eventhrough they plan to build/test on the specific ARCH. Here
we are not true whether community has a good way to resolve this. Here just
a suggest from us, how about describe the current testing status(all test
status) of Amplab in some place of spark? Then users can know spark testing
which already be pretested with generic test cases in upstream community,
and feel confident to use spark on anyplace what they want according to
that information.
In the end, we always believe community and follow community suggestions.
Please feel free to tell us about our outcomes at work, and welcome to work
together on Spark ARM. If any issue hit on ARM, please also @us for
discuss. ;-)
Thank you @Sean Owen <sr...@gmail.com> and team.
BR
ZhaoBo
[image: Mailtrack]
<https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&>
Sender
notified by
Mailtrack
<https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&>
19/11/19
下午05:49:00
Sean Owen <sr...@apache.org> 于2019年11月18日周一 上午10:06写道：
> Same response as before:
>
> - It is in the list of resolved JIRAs, of course
> - It (largely) worked previously
> - I think you're also saying you don't have 100% tests passing anyway,
> though probably just small issues
> - It does not seem to merit a special announcement from the PMC among
> the 2000+ changes in Spark 3
> - You are welcome to announce (on the project's user@ list if you
> like) whatever you want. Obviously, this is already well advertised on
> dev@
>
> I think you are asking for what borders on endorsement, and no that
> doesn't sound appropriate. Please just announce whatever you like as
> suggested.
>
> Sean
>
> On Sun, Nov 17, 2019 at 8:01 PM Tianhua huang <hu...@gmail.com>
> wrote:
> >
> > @Sean Owen，
> > I'm afraid I don't agree with you this time, I still remember no one can
> tell me whether Spark supports ARM or how much Spark can support ARM when I
> asked this first time on Dev@, you're very kind and told me to build and
> test on ARM locally and so sorry I think you were not sure much about this
> at that moment, right? Then I and my team work with community, we
> found/fixed several issues, integrate arm jobs into AMPLAB Jenkins, and the
> daily jobs has been stablely running for few weeks... after these efforts
> why not announce this officially in Spark releasenote? I believe after this
> everyone will know Spark is fully testing on ARM on community CI, Spark
> supports ARM basically, it's amazing and this will be very helpful. So what
> do you think? Or what are you worrying about?
>
Re: Ask for ARM CI for spark
Posted by Sean Owen <sr...@apache.org>.
Same response as before:
- It is in the list of resolved JIRAs, of course
- It (largely) worked previously
- I think you're also saying you don't have 100% tests passing anyway,
though probably just small issues
- It does not seem to merit a special announcement from the PMC among
the 2000+ changes in Spark 3
- You are welcome to announce (on the project's user@ list if you
like) whatever you want. Obviously, this is already well advertised on
dev@
I think you are asking for what borders on endorsement, and no that
doesn't sound appropriate. Please just announce whatever you like as
suggested.
Sean
On Sun, Nov 17, 2019 at 8:01 PM Tianhua huang <hu...@gmail.com> wrote:
>
> @Sean Owen，
> I'm afraid I don't agree with you this time, I still remember no one can tell me whether Spark supports ARM or how much Spark can support ARM when I asked this first time on Dev@, you're very kind and told me to build and test on ARM locally and so sorry I think you were not sure much about this at that moment, right? Then I and my team work with community, we found/fixed several issues, integrate arm jobs into AMPLAB Jenkins, and the daily jobs has been stablely running for few weeks... after these efforts why not announce this officially in Spark releasenote? I believe after this everyone will know Spark is fully testing on ARM on community CI, Spark supports ARM basically, it's amazing and this will be very helpful. So what do you think? Or what are you worrying about?
---------------------------------------------------------------------
To unsubscribe e-mail: dev-unsubscribe@spark.apache.org
Re: Ask for ARM CI for spark
Posted by Tianhua huang <hu...@gmail.com>.
@Sean Owen <sr...@gmail.com>，
I'm afraid I don't agree with you this time, I still remember no one can
tell me whether Spark supports ARM or how much Spark can support ARM when I
asked this first time on Dev@, you're very kind and told me to build and
test on ARM locally and so sorry I think you were not sure much about this
at that moment, right? Then I and my team work with community, we
found/fixed several issues, integrate arm jobs into AMPLAB Jenkins, and the
daily jobs has been stablely running for few weeks... after these efforts
why not announce this officially in Spark releasenote? I believe after this
everyone will know Spark is fully testing on ARM on community CI, Spark
supports ARM basically, it's amazing and this will be very helpful. So what
do you think? Or what are you worrying about?
On Mon, Nov 18, 2019 at 2:28 AM Steve Loughran <st...@cloudera.com> wrote:
> The ASF PR team would like something like that "Spark now supports ARM" in
> press releases. And don't forget: they do you like to be involved in the
> launch of the final release.
>
> On Fri, Nov 15, 2019 at 9:46 AM bo zhaobo <bz...@gmail.com>
> wrote:
>
>> Hi @Sean Owen <sr...@gmail.com> ,
>>
>> Thanks for your idea.
>>
>> We may use the bad words to describe our request. That's true that we
>> cannot just say "Spark support ARM from release 3.0.0", and we also cannot
>> say the past releases cannot run on ARM. But the reality is the past
>> releases didn't get a fully test on ARM like the current testing we do. And
>> that's true that current CI system have no resources can fit this kind
>> request(test on ARM).
>>
>> And please try to think, if a user wants to run lastest Spark release on
>> ARM(even the old releases), but community doesn't say that the specific
>> Spark release get testing on ARM. I think the users might think there is a
>> risk run on ARM, if he/she has no choice, they have to run spark on ARM,
>> they will build the CI system by themselves. That's very expensive. Right?
>> But now, community will do the same testing on ARM in the upstream, this
>> will save the users' resources. That's the reason announcing by community
>> in some ways is official and the best. Such as "In XXX release, Spark gets
>> fully testing on ARM" or "In XXX release, Spark community integrated an ARM
>> CI system. ". Once user see that, he/she would be very comfortable to use
>> Spark on ARM. ;-)
>>
>> Thanks for your paitent, we just discuss here, if I do something not
>> good, please feel free to correct and discuss. ;-)
>>
>> Thanks,
>>
>> BR
>>
>> ZhaoBo
>>
>>
>>
>>
>> [image: Mailtrack]
>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>> notified by
>> Mailtrack
>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/11/15
>> 下午05:43:57
>>
>> Sean Owen <sr...@gmail.com> 于2019年11月15日周五 下午5:04写道：
>>
>>> I don't think that's true either, not yet. Being JVM-based with no
>>> native code, I just don't even think it would be common to assume it
>>> doesn't work and it apparently has. If you want to announce it, that's
>>> up to you.
>>>
>>> On Fri, Nov 15, 2019 at 3:01 AM Tianhua huang <hu...@gmail.com>
>>> wrote:
>>> >
>>> > @Sean Owen,
>>> > Thanks for attention this.
>>> > I agree with you, it's probably not very appropriate to say 'support
>>> arm from 3.0 release'. How about change to the word "Spark community
>>> supports fully tests on arm from 3.0 release"?
>>> > Let's try to think about it from the user's point of view than
>>> developer，users have to know exactly whether spark supports arm well and
>>> wheter spark fully tests on arm. If we specify spark is fully tests on arm,
>>> I believe users will have much more confidence to run spark on arm.
>>> >
>>>
>>
Re: Ask for ARM CI for spark
Posted by Steve Loughran <st...@cloudera.com.INVALID>.
The ASF PR team would like something like that "Spark now supports ARM" in
press releases. And don't forget: they do you like to be involved in the
launch of the final release.
On Fri, Nov 15, 2019 at 9:46 AM bo zhaobo <bz...@gmail.com>
wrote:
> Hi @Sean Owen <sr...@gmail.com> ,
>
> Thanks for your idea.
>
> We may use the bad words to describe our request. That's true that we
> cannot just say "Spark support ARM from release 3.0.0", and we also cannot
> say the past releases cannot run on ARM. But the reality is the past
> releases didn't get a fully test on ARM like the current testing we do. And
> that's true that current CI system have no resources can fit this kind
> request(test on ARM).
>
> And please try to think, if a user wants to run lastest Spark release on
> ARM(even the old releases), but community doesn't say that the specific
> Spark release get testing on ARM. I think the users might think there is a
> risk run on ARM, if he/she has no choice, they have to run spark on ARM,
> they will build the CI system by themselves. That's very expensive. Right?
> But now, community will do the same testing on ARM in the upstream, this
> will save the users' resources. That's the reason announcing by community
> in some ways is official and the best. Such as "In XXX release, Spark gets
> fully testing on ARM" or "In XXX release, Spark community integrated an ARM
> CI system. ". Once user see that, he/she would be very comfortable to use
> Spark on ARM. ;-)
>
> Thanks for your paitent, we just discuss here, if I do something not good,
> please feel free to correct and discuss. ;-)
>
> Thanks,
>
> BR
>
> ZhaoBo
>
>
>
>
> [image: Mailtrack]
> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
> notified by
> Mailtrack
> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/11/15
> 下午05:43:57
>
> Sean Owen <sr...@gmail.com> 于2019年11月15日周五 下午5:04写道：
>
>> I don't think that's true either, not yet. Being JVM-based with no
>> native code, I just don't even think it would be common to assume it
>> doesn't work and it apparently has. If you want to announce it, that's
>> up to you.
>>
>> On Fri, Nov 15, 2019 at 3:01 AM Tianhua huang <hu...@gmail.com>
>> wrote:
>> >
>> > @Sean Owen,
>> > Thanks for attention this.
>> > I agree with you, it's probably not very appropriate to say 'support
>> arm from 3.0 release'. How about change to the word "Spark community
>> supports fully tests on arm from 3.0 release"?
>> > Let's try to think about it from the user's point of view than
>> developer，users have to know exactly whether spark supports arm well and
>> wheter spark fully tests on arm. If we specify spark is fully tests on arm,
>> I believe users will have much more confidence to run spark on arm.
>> >
>>
>
Re: Ask for ARM CI for spark
Posted by bo zhaobo <bz...@gmail.com>.
Hi @Sean Owen <sr...@gmail.com> ,
Thanks for your idea.
We may use the bad words to describe our request. That's true that we
cannot just say "Spark support ARM from release 3.0.0", and we also cannot
say the past releases cannot run on ARM. But the reality is the past
releases didn't get a fully test on ARM like the current testing we do. And
that's true that current CI system have no resources can fit this kind
request(test on ARM).
And please try to think, if a user wants to run lastest Spark release on
ARM(even the old releases), but community doesn't say that the specific
Spark release get testing on ARM. I think the users might think there is a
risk run on ARM, if he/she has no choice, they have to run spark on ARM,
they will build the CI system by themselves. That's very expensive. Right?
But now, community will do the same testing on ARM in the upstream, this
will save the users' resources. That's the reason announcing by community
in some ways is official and the best. Such as "In XXX release, Spark gets
fully testing on ARM" or "In XXX release, Spark community integrated an ARM
CI system. ". Once user see that, he/she would be very comfortable to use
Spark on ARM. ;-)
Thanks for your paitent, we just discuss here, if I do something not good,
please feel free to correct and discuss. ;-)
Thanks,
BR
ZhaoBo
[image: Mailtrack]
<https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&>
Sender
notified by
Mailtrack
<https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&>
19/11/15
下午05:43:57
Sean Owen <sr...@gmail.com> 于2019年11月15日周五 下午5:04写道：
> I don't think that's true either, not yet. Being JVM-based with no
> native code, I just don't even think it would be common to assume it
> doesn't work and it apparently has. If you want to announce it, that's
> up to you.
>
> On Fri, Nov 15, 2019 at 3:01 AM Tianhua huang <hu...@gmail.com>
> wrote:
> >
> > @Sean Owen,
> > Thanks for attention this.
> > I agree with you, it's probably not very appropriate to say 'support arm
> from 3.0 release'. How about change to the word "Spark community supports
> fully tests on arm from 3.0 release"?
> > Let's try to think about it from the user's point of view than
> developer，users have to know exactly whether spark supports arm well and
> wheter spark fully tests on arm. If we specify spark is fully tests on arm,
> I believe users will have much more confidence to run spark on arm.
> >
>
Re: Ask for ARM CI for spark
Posted by Sean Owen <sr...@gmail.com>.
I don't think that's true either, not yet. Being JVM-based with no
native code, I just don't even think it would be common to assume it
doesn't work and it apparently has. If you want to announce it, that's
up to you.
On Fri, Nov 15, 2019 at 3:01 AM Tianhua huang <hu...@gmail.com> wrote:
>
> @Sean Owen,
> Thanks for attention this.
> I agree with you, it's probably not very appropriate to say 'support arm from 3.0 release'. How about change to the word "Spark community supports fully tests on arm from 3.0 release"?
> Let's try to think about it from the user's point of view than developer，users have to know exactly whether spark supports arm well and wheter spark fully tests on arm. If we specify spark is fully tests on arm, I believe users will have much more confidence to run spark on arm.
>
---------------------------------------------------------------------
To unsubscribe e-mail: dev-unsubscribe@spark.apache.org
Re: Ask for ARM CI for spark
Posted by bo zhaobo <bz...@gmail.com>.
Hi,
And I found Spark-3.0.0-preview had released, but there is no releasenotes
in [1]. So how about to add support ARM notes in the next
releasenotes(maybe the releasenotes of Spark-3.0.0-preview). So sorry to
raise this, I'm not familiar with this, if any bad from me please feel free
to correct.
Moreover, since the daily job has been stablely running for few weeks,
probably we can say that the release have some basic ARM support in the
next release note.
From OpenSource project and Open Source eco system, this is a good chance
to tell peoples that Spark support ARM and support cross platforms. As
Spark can support more ARCH platforms, this is good to attract more users
to use Spark, eventhough they cannot use X86 in other uncontrollable
reasons.
[1] https://github.com/apache/spark-website/tree/asf-site/releases/_posts
[image: Mailtrack]
<https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&>
Sender
notified by
Mailtrack
<https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&>
19/11/15
上午11:16:28
bo zhaobo <bz...@gmail.com> 于2019年11月15日周五 上午11:00写道：
> Hi @Sean Owen <sr...@gmail.com> ,
>
> Thanks for reply. We know that Spark community has own release date and
> plan. We are happy to follow Spark community. But we think it's great if
> community could add a sentence into the next releasenotes and claim "Spark
> can support Arm from this release." after we finish the test work on ARM.
> That's all. We just want a community official caliber that spark support
> ARM for attracting more users to use spark.
>
> Thank you very much for your patient.
>
> BR
>
> ZhaoBo
>
>
> [image: Mailtrack]
> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
> notified by
> Mailtrack
> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/11/15
> 上午10:59:29
>
> Tianhua huang <hu...@gmail.com> 于2019年11月15日周五 上午10:25写道：
>
>> @Sean,
>> Yes, you are right, we don't have to create a separate release of Spark
>> for ARM, it's enough to add a releasenote to say that Spark supports
>> arm architecture.
>> About the test failure, one or two tests will timeout on our poor
>> performance arm instance sometimes, now we donate a high performance arm
>> instance to amplab, and wait shane to build the jobs on it.
>>
>> On Fri, Nov 15, 2019 at 10:13 AM Sean Owen <sr...@gmail.com> wrote:
>>
>>> I don't quite understand. You are saying tests don't pass yet, so why
>>> would anyone yet run these tests regularly?
>>> If it's because the instances aren't fast enough, use bigger instances?
>>> I don't think anyone would create a separate release of Spark for ARM,
>>> no. But why would that be necessary?
>>>
>>> On Thu, Nov 14, 2019 at 7:28 PM bo zhaobo <bz...@gmail.com>
>>> wrote:
>>>
>>>> Hi Spark team,
>>>>
>>>> Any ideas about the above email? Thank you.
>>>>
>>>> BR
>>>>
>>>> ZhaoBo
>>>>
>>>>
>>>> [image: Mailtrack]
>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>> notified by
>>>> Mailtrack
>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/11/15
>>>> 上午09:26:17
>>>>
>>>> Tianhua huang <hu...@gmail.com> 于2019年11月12日周二 下午2:47写道：
>>>>
>>>>> Hi all,
>>>>>
>>>>> Spark arm jobs have built for some time, and now there are two jobs[1]
>>>>> spark-master-test-maven-arm
>>>>> <https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-maven-arm/>
>>>>> and spark-master-test-python-arm
>>>>> <https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-python-arm/>,
>>>>> we can see there are some build failures, but it because of the poor
>>>>> performance of the arm instance, and now we begin to build spark arm jobs
>>>>> on other high performance instances, and the build/test are all success, we
>>>>> plan to donate the instance to amplab later. According to the build
>>>>> history, we are very happy to say spark is supported on aarch64 platform,
>>>>> and I suggest to add this good news into spark-3.0.0 releasenotes. Maybe
>>>>> community could provide an arm-supported release of spark at the meanwhile?
>>>>>
>>>>> [1]
>>>>> https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-maven-arm/
>>>>>
>>>>>
>>>>> https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-python-arm/
>>>>>
>>>>>
>>>>> ps: the jira https://issues.apache.org/jira/browse/SPARK-29106 trace
>>>>> the whole work, thank you very much Shane:)
>>>>>
>>>>
Re: Ask for ARM CI for spark
Posted by Tianhua huang <hu...@gmail.com>.
@Sean Owen,
Thanks for attention this.
I agree with you, it's probably not very appropriate to say 'support arm
from 3.0 release'. How about change to the word "Spark community supports
fully tests on arm from 3.0 release"?
Let's try to think about it from the user's point of view than
developer，users have to know exactly whether spark supports arm well and
wheter spark fully tests on arm. If we specify spark is fully tests on arm,
I believe users will have much more confidence to run spark on arm.
On Fri, Nov 15, 2019 at 4:05 PM Sean Owen <sr...@gmail.com> wrote:
> I'm not against it, but the JIRAs will already show that the small
> ARM-related difference like floating-point in log() were resolved.
> Those aren't major enough to highlight as key changes in the 2000+
> resolved. it didn't really not-work before either, as I understand;
> Spark isn't specific to an architecture, so I don't know if that
> situation changed materially in 3.0; it still otherwise ran in 2.x on
> ARM right? It would imply people couldn't use it on ARM previously.
> You can certainly announce you endorse 3.0 as a good release for ARM
> and/or call attention to it on user@.
>
> On Thu, Nov 14, 2019 at 9:01 PM bo zhaobo <bz...@gmail.com>
> wrote:
> >
> > Hi @Sean Owen ,
> >
> > Thanks for reply. We know that Spark community has own release date and
> plan. We are happy to follow Spark community. But we think it's great if
> community could add a sentence into the next releasenotes and claim "Spark
> can support Arm from this release." after we finish the test work on ARM.
> That's all. We just want a community official caliber that spark support
> ARM for attracting more users to use spark.
> >
>
Re: Ask for ARM CI for spark
Posted by Sean Owen <sr...@gmail.com>.
I'm not against it, but the JIRAs will already show that the small
ARM-related difference like floating-point in log() were resolved.
Those aren't major enough to highlight as key changes in the 2000+
resolved. it didn't really not-work before either, as I understand;
Spark isn't specific to an architecture, so I don't know if that
situation changed materially in 3.0; it still otherwise ran in 2.x on
ARM right? It would imply people couldn't use it on ARM previously.
You can certainly announce you endorse 3.0 as a good release for ARM
and/or call attention to it on user@.
On Thu, Nov 14, 2019 at 9:01 PM bo zhaobo <bz...@gmail.com> wrote:
>
> Hi @Sean Owen ,
>
> Thanks for reply. We know that Spark community has own release date and plan. We are happy to follow Spark community. But we think it's great if community could add a sentence into the next releasenotes and claim "Spark can support Arm from this release." after we finish the test work on ARM. That's all. We just want a community official caliber that spark support ARM for attracting more users to use spark.
>
---------------------------------------------------------------------
To unsubscribe e-mail: dev-unsubscribe@spark.apache.org
Re: Ask for ARM CI for spark
Posted by bo zhaobo <bz...@gmail.com>.
Hi @Sean Owen <sr...@gmail.com> ,
Thanks for reply. We know that Spark community has own release date and
plan. We are happy to follow Spark community. But we think it's great if
community could add a sentence into the next releasenotes and claim "Spark
can support Arm from this release." after we finish the test work on ARM.
That's all. We just want a community official caliber that spark support
ARM for attracting more users to use spark.
Thank you very much for your patient.
BR
ZhaoBo
[image: Mailtrack]
<https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&>
Sender
notified by
Mailtrack
<https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&>
19/11/15
上午10:59:29
Tianhua huang <hu...@gmail.com> 于2019年11月15日周五 上午10:25写道：
> @Sean,
> Yes, you are right, we don't have to create a separate release of Spark
> for ARM, it's enough to add a releasenote to say that Spark supports
> arm architecture.
> About the test failure, one or two tests will timeout on our poor
> performance arm instance sometimes, now we donate a high performance arm
> instance to amplab, and wait shane to build the jobs on it.
>
> On Fri, Nov 15, 2019 at 10:13 AM Sean Owen <sr...@gmail.com> wrote:
>
>> I don't quite understand. You are saying tests don't pass yet, so why
>> would anyone yet run these tests regularly?
>> If it's because the instances aren't fast enough, use bigger instances?
>> I don't think anyone would create a separate release of Spark for ARM,
>> no. But why would that be necessary?
>>
>> On Thu, Nov 14, 2019 at 7:28 PM bo zhaobo <bz...@gmail.com>
>> wrote:
>>
>>> Hi Spark team,
>>>
>>> Any ideas about the above email? Thank you.
>>>
>>> BR
>>>
>>> ZhaoBo
>>>
>>>
>>> [image: Mailtrack]
>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>> notified by
>>> Mailtrack
>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/11/15
>>> 上午09:26:17
>>>
>>> Tianhua huang <hu...@gmail.com> 于2019年11月12日周二 下午2:47写道：
>>>
>>>> Hi all,
>>>>
>>>> Spark arm jobs have built for some time, and now there are two jobs[1]
>>>> spark-master-test-maven-arm
>>>> <https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-maven-arm/>
>>>> and spark-master-test-python-arm
>>>> <https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-python-arm/>,
>>>> we can see there are some build failures, but it because of the poor
>>>> performance of the arm instance, and now we begin to build spark arm jobs
>>>> on other high performance instances, and the build/test are all success, we
>>>> plan to donate the instance to amplab later. According to the build
>>>> history, we are very happy to say spark is supported on aarch64 platform,
>>>> and I suggest to add this good news into spark-3.0.0 releasenotes. Maybe
>>>> community could provide an arm-supported release of spark at the meanwhile?
>>>>
>>>> [1]
>>>> https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-maven-arm/
>>>>
>>>> https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-python-arm/
>>>>
>>>>
>>>> ps: the jira https://issues.apache.org/jira/browse/SPARK-29106 trace
>>>> the whole work, thank you very much Shane:)
>>>>
>>>
Re: Ask for ARM CI for spark
Posted by Tianhua huang <hu...@gmail.com>.
@Sean,
Yes, you are right, we don't have to create a separate release of Spark for
ARM, it's enough to add a releasenote to say that Spark supports
arm architecture.
About the test failure, one or two tests will timeout on our poor
performance arm instance sometimes, now we donate a high performance arm
instance to amplab, and wait shane to build the jobs on it.
On Fri, Nov 15, 2019 at 10:13 AM Sean Owen <sr...@gmail.com> wrote:
> I don't quite understand. You are saying tests don't pass yet, so why
> would anyone yet run these tests regularly?
> If it's because the instances aren't fast enough, use bigger instances?
> I don't think anyone would create a separate release of Spark for ARM, no.
> But why would that be necessary?
>
> On Thu, Nov 14, 2019 at 7:28 PM bo zhaobo <bz...@gmail.com>
> wrote:
>
>> Hi Spark team,
>>
>> Any ideas about the above email? Thank you.
>>
>> BR
>>
>> ZhaoBo
>>
>>
>> [image: Mailtrack]
>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>> notified by
>> Mailtrack
>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/11/15
>> 上午09:26:17
>>
>> Tianhua huang <hu...@gmail.com> 于2019年11月12日周二 下午2:47写道：
>>
>>> Hi all,
>>>
>>> Spark arm jobs have built for some time, and now there are two jobs[1]
>>> spark-master-test-maven-arm
>>> <https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-maven-arm/>
>>> and spark-master-test-python-arm
>>> <https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-python-arm/>,
>>> we can see there are some build failures, but it because of the poor
>>> performance of the arm instance, and now we begin to build spark arm jobs
>>> on other high performance instances, and the build/test are all success, we
>>> plan to donate the instance to amplab later. According to the build
>>> history, we are very happy to say spark is supported on aarch64 platform,
>>> and I suggest to add this good news into spark-3.0.0 releasenotes. Maybe
>>> community could provide an arm-supported release of spark at the meanwhile?
>>>
>>> [1]
>>> https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-maven-arm/
>>> https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-python-arm/
>>>
>>>
>>> ps: the jira https://issues.apache.org/jira/browse/SPARK-29106 trace
>>> the whole work, thank you very much Shane:)
>>>
>>
Re: Ask for ARM CI for spark
Posted by Sean Owen <sr...@gmail.com>.
I don't quite understand. You are saying tests don't pass yet, so why would
anyone yet run these tests regularly?
If it's because the instances aren't fast enough, use bigger instances?
I don't think anyone would create a separate release of Spark for ARM, no.
But why would that be necessary?
On Thu, Nov 14, 2019 at 7:28 PM bo zhaobo <bz...@gmail.com>
wrote:
> Hi Spark team,
>
> Any ideas about the above email? Thank you.
>
> BR
>
> ZhaoBo
>
>
> [image: Mailtrack]
> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
> notified by
> Mailtrack
> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/11/15
> 上午09:26:17
>
> Tianhua huang <hu...@gmail.com> 于2019年11月12日周二 下午2:47写道：
>
>> Hi all,
>>
>> Spark arm jobs have built for some time, and now there are two jobs[1]
>> spark-master-test-maven-arm
>> <https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-maven-arm/>
>> and spark-master-test-python-arm
>> <https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-python-arm/>,
>> we can see there are some build failures, but it because of the poor
>> performance of the arm instance, and now we begin to build spark arm jobs
>> on other high performance instances, and the build/test are all success, we
>> plan to donate the instance to amplab later. According to the build
>> history, we are very happy to say spark is supported on aarch64 platform,
>> and I suggest to add this good news into spark-3.0.0 releasenotes. Maybe
>> community could provide an arm-supported release of spark at the meanwhile?
>>
>> [1]
>> https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-maven-arm/
>> https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-python-arm/
>>
>> ps: the jira https://issues.apache.org/jira/browse/SPARK-29106 trace the
>> whole work, thank you very much Shane:)
>>
>
Re: Ask for ARM CI for spark
Posted by bo zhaobo <bz...@gmail.com>.
Hi Spark team,
Any ideas about the above email? Thank you.
BR
ZhaoBo
[image: Mailtrack]
<https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&>
Sender
notified by
Mailtrack
<https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&>
19/11/15
上午09:26:17
Tianhua huang <hu...@gmail.com> 于2019年11月12日周二 下午2:47写道：
> Hi all,
>
> Spark arm jobs have built for some time, and now there are two jobs[1]
> spark-master-test-maven-arm
> <https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-maven-arm/>
> and spark-master-test-python-arm
> <https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-python-arm/>,
> we can see there are some build failures, but it because of the poor
> performance of the arm instance, and now we begin to build spark arm jobs
> on other high performance instances, and the build/test are all success, we
> plan to donate the instance to amplab later. According to the build
> history, we are very happy to say spark is supported on aarch64 platform,
> and I suggest to add this good news into spark-3.0.0 releasenotes. Maybe
> community could provide an arm-supported release of spark at the meanwhile?
>
> [1]
> https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-maven-arm/
> https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-python-arm/
>
> ps: the jira https://issues.apache.org/jira/browse/SPARK-29106 trace the
> whole work, thank you very much Shane:)
>
> On Thu, Oct 17, 2019 at 2:52 PM bo zhaobo <bz...@gmail.com>
> wrote:
>
>> Just Notes: The jira issue link is
>> https://issues.apache.org/jira/browse/SPARK-29106
>>
>>
>>
>> [image: Mailtrack]
>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>> notified by
>> Mailtrack
>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/10/17
>> 下午02:50:01
>>
>> Tianhua huang <hu...@gmail.com> 于2019年10月17日周四 上午10:47写道：
>>
>>> OK, let's update infos there. Thanks.
>>>
>>> On Thu, Oct 17, 2019 at 1:52 AM Shane Knapp <sk...@berkeley.edu> wrote:
>>>
>>>> i totally missed the spark jira from earlier... let's move the
>>>> conversation there!
>>>>
>>>> On Tue, Oct 15, 2019 at 6:21 PM bo zhaobo <bz...@gmail.com>
>>>> wrote:
>>>>
>>>>> Shane, Awaresome! We will try the best to finish the test and the
>>>>> requests on the VM recently. Once we finish those things, we will send you
>>>>> an email , then we can continue the following things. Thank you very much.
>>>>>
>>>>> Best Regards,
>>>>>
>>>>> ZhaoBo
>>>>>
>>>>> Shane Knapp <sk...@berkeley.edu> 于 2019年10月16日周三 上午3:47写道：
>>>>>
>>>>>> ok! i'm able to successfully log in to the VM!
>>>>>>
>>>>>> i also have created a jenkins worker entry:
>>>>>> https://amplab.cs.berkeley.edu/jenkins/computer/spark-arm-vm/
>>>>>>
>>>>>> it's a pretty bare-bones VM, so i have some suggestions/requests
>>>>>> before we can actually proceed w/testing. i will not be able to perform
>>>>>> any system configuration, as i don't have the cycles to reverse-engineer
>>>>>> the ansible setup and test it all out.
>>>>>>
>>>>>> * java is not installed, please install the following:
>>>>>> - java8 min version 1.8.0_191
>>>>>> - java11 min version 11.0.1
>>>>>>
>>>>>> * it appears from the ansible playbook that there are other deps that
>>>>>> need to be installed.
>>>>>> - please install all deps
>>>>>> - manually run the tests until they pass
>>>>>>
>>>>>> * the jenkins user should NEVER have sudo or any root-level access!
>>>>>>
>>>>>> * once the arm tests pass when manually run, take a snapshot of this
>>>>>> image so we can recreate it w/o needing to reinstall everything
>>>>>>
>>>>>> after that's done i can finish configuring the jenkins worker and set
>>>>>> up a build...
>>>>>>
>>>>>> thanks!
>>>>>>
>>>>>> shane
>>>>>>
>>>>>>
>>>>>> On Mon, Oct 14, 2019 at 8:34 PM Shane Knapp <sk...@berkeley.edu>
>>>>>> wrote:
>>>>>>
>>>>>>> yes, i will get to that tomorrow. today was spent cleaning up the
>>>>>>> mess from last week.
>>>>>>>
>>>>>>> On Mon, Oct 14, 2019 at 6:18 PM bo zhaobo <
>>>>>>> bzhaojyathousandy@gmail.com> wrote:
>>>>>>>
>>>>>>>> Hi shane,
>>>>>>>>
>>>>>>>> That's great news for Amplab is back. ;-) . If possible, could you
>>>>>>>> please take several minutes to check the ARM VM is accessible from your
>>>>>>>> side? And is there any plan for the whole ARM test integration from
>>>>>>>> you?(how about we finish it this month?) Thanks.
>>>>>>>>
>>>>>>>> Best regards,
>>>>>>>>
>>>>>>>> ZhaoBo
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> [image: Mailtrack]
>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>>>> notified by
>>>>>>>> Mailtrack
>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/10/15
>>>>>>>> 上午09:13:33
>>>>>>>>
>>>>>>>> bo zhaobo <bz...@gmail.com> 于2019年10月10日周四 上午8:29写道：
>>>>>>>>
>>>>>>>>> Oh, sorry about we miss that email. If possible, could you please
>>>>>>>>> take some minutes to test the ARM VM is accessible through your ssh private
>>>>>>>>> key with jenkins user? And we plan to make the whole integration process
>>>>>>>>> and test could be done before the end of this month. We are very happy to
>>>>>>>>> work together with you to move it forward, if you are free and agree that.
>>>>>>>>> :) Thank you very much.
>>>>>>>>>
>>>>>>>>> Best Regards,
>>>>>>>>> Zhao Bo
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Shane Knapp <sk...@berkeley.edu> 于 2019年10月9日周三 下午11:10写道：
>>>>>>>>>
>>>>>>>>>> i spent yesterday dealing w/a power outage on campus. please see
>>>>>>>>>> my email to the spark dev list.
>>>>>>>>>>
>>>>>>>>>> On Wed, Oct 9, 2019 at 3:29 AM Tianhua huang <
>>>>>>>>>> huangtianhua223@gmail.com> wrote:
>>>>>>>>>>
>>>>>>>>>>> Hi Shane,
>>>>>>>>>>> Sorry to disturb you again, I wonder if there is a progress of
>>>>>>>>>>> the jenkins arm job?
>>>>>>>>>>> And please don't hesitate to contact us if you have any
>>>>>>>>>>> questions and need helps, thank you very much :)
>>>>>>>>>>>
>>>>>>>>>>> On Tue, Oct 8, 2019 at 9:21 AM bo zhaobo <
>>>>>>>>>>> bzhaojyathousandy@gmail.com> wrote:
>>>>>>>>>>>
>>>>>>>>>>>> Hi Shane,
>>>>>>>>>>>>
>>>>>>>>>>>> Sorry for late. We are just back for a 7-days holiday.
>>>>>>>>>>>>
>>>>>>>>>>>> I already make the public key into the VM. If you are free,
>>>>>>>>>>>> please go ahead to test. Thank you
>>>>>>>>>>>>
>>>>>>>>>>>> Best regards
>>>>>>>>>>>>
>>>>>>>>>>>> ZhaoBo
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>> [image: Mailtrack]
>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>>>>>>>> notified by
>>>>>>>>>>>> Mailtrack
>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/10/08
>>>>>>>>>>>> 上午09:16:49
>>>>>>>>>>>>
>>>>>>>>>>>> Shane Knapp <sk...@berkeley.edu> 于2019年10月1日周二 上午2:33写道：
>>>>>>>>>>>>
>>>>>>>>>>>>> here's the public key for the jenkins user:
>>>>>>>>>>>>> ssh-rsa
>>>>>>>>>>>>> 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
>>>>>>>>>>>>> jenkins@amp-jenkins-master
>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>> once you get that set up, please let me know and i will test
>>>>>>>>>>>>> ssh-ability.
>>>>>>>>>>>>>
>>>>>>>>>>>>> On Wed, Sep 25, 2019 at 10:15 PM Shane Knapp <
>>>>>>>>>>>>> sknapp@berkeley.edu> wrote:
>>>>>>>>>>>>>
>>>>>>>>>>>>>> thanks for understanding. i'm a one-man show. :)
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> shane
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> On Wed, Sep 25, 2019 at 6:52 PM Tianhua huang <
>>>>>>>>>>>>>> huangtianhua223@gmail.com> wrote:
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> I am very happy to hear that :)
>>>>>>>>>>>>>>> And please don't hesitate to contact us if you have any
>>>>>>>>>>>>>>> questions and need helps, thanks again.
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> On Thu, Sep 26, 2019 at 12:34 AM Shane Knapp <
>>>>>>>>>>>>>>> sknapp@berkeley.edu> wrote:
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> i'm hoping to get started by EOW or beginning of next. i
>>>>>>>>>>>>>>>> have an incredibly small team here @ berkeley and support a lot of research
>>>>>>>>>>>>>>>> labs, and i've been down 25-50% staffing for the past 6 weeks. thankfully
>>>>>>>>>>>>>>>> my whole team is back and i'm finally getting my head above water and will
>>>>>>>>>>>>>>>> have time to dedicate to this really soon.
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> sorry for the delay!
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> shane
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> On Mon, Sep 23, 2019 at 7:21 PM bo zhaobo <
>>>>>>>>>>>>>>>> bzhaojyathousandy@gmail.com> wrote:
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>> Hi Shane,
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>> How about to start the ARM work this week? Can we? ;-).
>>>>>>>>>>>>>>>>> Thank you
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>> Best Regards
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>> [image: Mailtrack]
>>>>>>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>>>>>>>>>>>>> notified by
>>>>>>>>>>>>>>>>> Mailtrack
>>>>>>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/09/24
>>>>>>>>>>>>>>>>> 上午10:18:01
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>> Shane Knapp <sk...@berkeley.edu> 于2019年9月20日周五 上午11:57写道：
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>> i'll have the cycles over the next week.
>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>> On Thu, Sep 19, 2019 at 7:51 PM bo zhaobo <
>>>>>>>>>>>>>>>>>> bzhaojyathousandy@gmail.com> wrote:
>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>> Hi Shane,
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>> Is there any update about the last email? ;-)
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>> Thanks,
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>> Best Regards
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>> [image: Mailtrack]
>>>>>>>>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>>>>>>>>>>>>>>> notified by
>>>>>>>>>>>>>>>>>>> Mailtrack
>>>>>>>>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/09/20
>>>>>>>>>>>>>>>>>>> 上午10:49:35
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>> bo zhaobo <bz...@gmail.com> 于2019年9月18日周三
>>>>>>>>>>>>>>>>>>> 上午10:02写道：
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>> Thanks for reply.
>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>> I will answer you question One by One.
>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>> 0) Sure, I can create an user named 'jenkins' manually.
>>>>>>>>>>>>>>>>>>>> And I will left the 'arm' User if you might need the higher priority to do
>>>>>>>>>>>>>>>>>>>> something on the VM.
>>>>>>>>>>>>>>>>>>>> 1) Sure, that would be great if you could provide a
>>>>>>>>>>>>>>>>>>>> public key to us. ;-)
>>>>>>>>>>>>>>>>>>>> 2) Sure
>>>>>>>>>>>>>>>>>>>> 3) Yeah, it's a persistent VM now. But from us, we plan
>>>>>>>>>>>>>>>>>>>> to donate more resources beside the end of October. It's better to move the
>>>>>>>>>>>>>>>>>>>> all ARM jenkins workers to the same place, so this VM might be recycle in
>>>>>>>>>>>>>>>>>>>> the future. For now, that dosen't stop us to move the CI job forward.
>>>>>>>>>>>>>>>>>>>> 4) Correct. ;-)
>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>> If you need more information or help about the ARM VM,
>>>>>>>>>>>>>>>>>>>> please free to contact us.
>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>> Questions from us:
>>>>>>>>>>>>>>>>>>>> 1> is there a online chat platform? then we can chat
>>>>>>>>>>>>>>>>>>>> and discuss on that. Seem emailing is too slow for us. ;-). Or is there a
>>>>>>>>>>>>>>>>>>>> online discuss platform in Spark community?
>>>>>>>>>>>>>>>>>>>> 2> Could you please share the test scripts once you
>>>>>>>>>>>>>>>>>>>> finish the test integration? Also the existing Jenkins test scripts in
>>>>>>>>>>>>>>>>>>>> Spark CI, as we plan to do the same thing like X86, so We think testing the
>>>>>>>>>>>>>>>>>>>> same jobs like X86 on ARM would be a good start.
>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>> Thank you
>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>> Best Regards
>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>> [image: Mailtrack]
>>>>>>>>>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>>>>>>>>>>>>>>>> notified by
>>>>>>>>>>>>>>>>>>>> Mailtrack
>>>>>>>>>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/09/18
>>>>>>>>>>>>>>>>>>>> 上午09:33:53
>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>> Shane Knapp <sk...@berkeley.edu> 于2019年9月18日周三
>>>>>>>>>>>>>>>>>>>> 上午5:05写道：
>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>> thanks for the info... a couple of things:
>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>> 0) could you rename/create a 'jenkins' user with the
>>>>>>>>>>>>>>>>>>>>> same rights?
>>>>>>>>>>>>>>>>>>>>> 1) i will provide an ssh key so that you can add it to
>>>>>>>>>>>>>>>>>>>>> the jenkins user's authorized_keys file
>>>>>>>>>>>>>>>>>>>>> 2) the jenkins user should NOT have root access. this
>>>>>>>>>>>>>>>>>>>>> is a major security hole
>>>>>>>>>>>>>>>>>>>>> 3) will this be a persistent VM? if so, i'd much
>>>>>>>>>>>>>>>>>>>>> prefer to have it set up initially so we can just log in, build what we
>>>>>>>>>>>>>>>>>>>>> need and launch the job
>>>>>>>>>>>>>>>>>>>>> 4) yay ansible! i can create the job locally from the
>>>>>>>>>>>>>>>>>>>>> template.
>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>> On Tue, Sep 17, 2019 at 1:16 AM bo zhaobo <
>>>>>>>>>>>>>>>>>>>>> bzhaojyathousandy@gmail.com> wrote:
>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>> Hi @shane knapp <sk...@berkeley.edu> , Thanks.
>>>>>>>>>>>>>>>>>>>>>> Tianhua huang and I already create a ARM VM for this.
>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>> The Infomation about the ARM VM:
>>>>>>>>>>>>>>>>>>>>>> IP: 213.146.141.16
>>>>>>>>>>>>>>>>>>>>>> ssh_key: please see the attachment file
>>>>>>>>>>>>>>>>>>>>>> "lab_ssh_key.txt"
>>>>>>>>>>>>>>>>>>>>>> Then you can login to it with "ssh -i
>>>>>>>>>>>>>>>>>>>>>> lab_ssh_key.txt arm@213.146.141.16 ", the User "arm"
>>>>>>>>>>>>>>>>>>>>>> have sudo priority.
>>>>>>>>>>>>>>>>>>>>>> Notes: This VM is running on Cloud, now it just
>>>>>>>>>>>>>>>>>>>>>> enable ssh (tcp 22 port) and ping (icmp) for ingress network connection, if
>>>>>>>>>>>>>>>>>>>>>> you need more, please let us know.
>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>> And first we can add periodic job for spark arm test,
>>>>>>>>>>>>>>>>>>>>>> maybe the zuul job details of openlab we tested can be referenced in
>>>>>>>>>>>>>>>>>>>>>> jenkins job
>>>>>>>>>>>>>>>>>>>>>> https://github.com/theopenlab/openlab-zuul-jobs/blob/master/playbooks/spark-unit-test-arm64/run.yaml
>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>> Let's give you some explaination about that file. In
>>>>>>>>>>>>>>>>>>>>>> the Step3 of above file, leveldbjni doesn't have a ARM release, so we build
>>>>>>>>>>>>>>>>>>>>>> leveldbjni on arm and use ourselves leveldbjni jar, which locates on [1].
>>>>>>>>>>>>>>>>>>>>>> So you need to still use the jar to test. And hadoop side also depends on
>>>>>>>>>>>>>>>>>>>>>> leveldbjni, I remember that is "hadoop-client". For this reason, we can not
>>>>>>>>>>>>>>>>>>>>>> change the pom file directly, so we use "mvn install" to enable leveldb
>>>>>>>>>>>>>>>>>>>>>> during testing spark on ARM.
>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>> Please Notes： All test runs without ROOT user. So
>>>>>>>>>>>>>>>>>>>>>> that's good if you test with "arm" User.
>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>> Thanks,
>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>> Best Regards
>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>> [1]
>>>>>>>>>>>>>>>>>>>>>> https://repo1.maven.org/maven2/org/openlabtesting/leveldbjni/leveldbjni-all/1.8/leveldbjni-all-1.8.jar
>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>> [image: Mailtrack]
>>>>>>>>>>>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>>>>>>>>>>>>>>>>>> notified by
>>>>>>>>>>>>>>>>>>>>>> Mailtrack
>>>>>>>>>>>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/09/17
>>>>>>>>>>>>>>>>>>>>>> 下午04:08:59
>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>> Tianhua huang <hu...@gmail.com>
>>>>>>>>>>>>>>>>>>>>>> 于2019年9月17日周二 下午2:48写道：
>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>> @shane knapp <sk...@berkeley.edu> thank you very
>>>>>>>>>>>>>>>>>>>>>>> much, I opened an issue for this
>>>>>>>>>>>>>>>>>>>>>>> https://issues.apache.org/jira/browse/SPARK-29106,
>>>>>>>>>>>>>>>>>>>>>>> we can tall the details in it :)
>>>>>>>>>>>>>>>>>>>>>>> And we will prepare an arm instance today and will
>>>>>>>>>>>>>>>>>>>>>>> send the info to your email later.
>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>> On Tue, Sep 17, 2019 at 4:40 AM Shane Knapp <
>>>>>>>>>>>>>>>>>>>>>>> sknapp@berkeley.edu> wrote:
>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>> @Tianhua huang <hu...@gmail.com> sure, i
>>>>>>>>>>>>>>>>>>>>>>>> think we can get something sorted for the short-term.
>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>> all we need is ssh access (i can provide an ssh
>>>>>>>>>>>>>>>>>>>>>>>> key), and i can then have our jenkins master launch a remote worker on that
>>>>>>>>>>>>>>>>>>>>>>>> instance.
>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>> instance setup, etc, will be up to you. my support
>>>>>>>>>>>>>>>>>>>>>>>> for the time being will be to create the job and 'best effort' for
>>>>>>>>>>>>>>>>>>>>>>>> everything else.
>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>> this should get us up and running asap.
>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>> is there an open JIRA for jenkins/arm test
>>>>>>>>>>>>>>>>>>>>>>>> support? we can move the technical details about this idea there.
>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>> On Sun, Sep 15, 2019 at 9:03 PM Tianhua huang <
>>>>>>>>>>>>>>>>>>>>>>>> huangtianhua223@gmail.com> wrote:
>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>> @Sean Owen <sr...@gmail.com> , so sorry to reply
>>>>>>>>>>>>>>>>>>>>>>>>> late, we had a Mid-Autumn holiday:)
>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>> If you hope to integrate ARM CI to amplab jenkins,
>>>>>>>>>>>>>>>>>>>>>>>>> we can offer the arm instance, and then the ARM job will run together with
>>>>>>>>>>>>>>>>>>>>>>>>> other x86 jobs, so maybe there is a guideline to do this? @shane
>>>>>>>>>>>>>>>>>>>>>>>>> knapp <sk...@berkeley.edu> would you help us?
>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>> On Thu, Sep 12, 2019 at 9:36 PM Sean Owen <
>>>>>>>>>>>>>>>>>>>>>>>>> srowen@gmail.com> wrote:
>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>> I don't know what's involved in actually
>>>>>>>>>>>>>>>>>>>>>>>>>> accepting or operating those machines, so can't comment there, but in the
>>>>>>>>>>>>>>>>>>>>>>>>>> meantime it's good that you are running these tests and can help report
>>>>>>>>>>>>>>>>>>>>>>>>>> changes needed to keep it working with ARM. I would continue with that for
>>>>>>>>>>>>>>>>>>>>>>>>>> now.
>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>> On Wed, Sep 11, 2019 at 10:06 PM Tianhua huang <
>>>>>>>>>>>>>>>>>>>>>>>>>> huangtianhua223@gmail.com> wrote:
>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>> Hi all,
>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>> For the whole work process of spark ARM CI, we
>>>>>>>>>>>>>>>>>>>>>>>>>>> want to make 2 things clear.
>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>> The first thing is:
>>>>>>>>>>>>>>>>>>>>>>>>>>> About spark ARM CI, now we have two periodic
>>>>>>>>>>>>>>>>>>>>>>>>>>> jobs, one job[1] based on commit[2](which already fixed the replay tests
>>>>>>>>>>>>>>>>>>>>>>>>>>> failed issue[3], we made a new test branch based on date 09-09-2019), the
>>>>>>>>>>>>>>>>>>>>>>>>>>> other job[4] based on spark master.
>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>> The first job we test on the specified branch to
>>>>>>>>>>>>>>>>>>>>>>>>>>> prove that our ARM CI is good and stable.
>>>>>>>>>>>>>>>>>>>>>>>>>>> The second job checks spark master every day,
>>>>>>>>>>>>>>>>>>>>>>>>>>> then we can find whether the latest commits affect the ARM CI. According to
>>>>>>>>>>>>>>>>>>>>>>>>>>> the build history and result, it shows that some problems are easier to
>>>>>>>>>>>>>>>>>>>>>>>>>>> find on ARM like SPARK-28770
>>>>>>>>>>>>>>>>>>>>>>>>>>> <https://issues.apache.org/jira/browse/SPARK-28770>,
>>>>>>>>>>>>>>>>>>>>>>>>>>> and it also shows that we would make efforts to trace and figure them
>>>>>>>>>>>>>>>>>>>>>>>>>>> out, till now we have found and fixed several problems[5][6][7], thanks
>>>>>>>>>>>>>>>>>>>>>>>>>>> everyone of the community :). And we believe that ARM CI is very necessary,
>>>>>>>>>>>>>>>>>>>>>>>>>>> right?
>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>> The second thing is:
>>>>>>>>>>>>>>>>>>>>>>>>>>> We plan to run the jobs for a period of time,
>>>>>>>>>>>>>>>>>>>>>>>>>>> and you can see the result and logs from 'build history' of the jobs
>>>>>>>>>>>>>>>>>>>>>>>>>>> console, if everything goes well for one or two weeks could community
>>>>>>>>>>>>>>>>>>>>>>>>>>> accept the ARM CI? or how long the periodic jobs to run then our community
>>>>>>>>>>>>>>>>>>>>>>>>>>> could have enough confidence to accept the ARM CI? As you suggested before,
>>>>>>>>>>>>>>>>>>>>>>>>>>> it's good to integrate ARM CI to amplab jenkins, we agree that and we can
>>>>>>>>>>>>>>>>>>>>>>>>>>> donate the ARM instances and then maintain the ARM-related test jobs
>>>>>>>>>>>>>>>>>>>>>>>>>>> together with community, any thoughts?
>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>> Thank you all!
>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>> [1]
>>>>>>>>>>>>>>>>>>>>>>>>>>> http://status.openlabtesting.org/job/spark-unchanged-branch-unit-test-hadoop-2.7-arm64
>>>>>>>>>>>>>>>>>>>>>>>>>>> [2]
>>>>>>>>>>>>>>>>>>>>>>>>>>> https://github.com/apache/spark/commit/0ed9fae45769d4b06b8cf8128f462f09ff3d9a72
>>>>>>>>>>>>>>>>>>>>>>>>>>> [3]
>>>>>>>>>>>>>>>>>>>>>>>>>>> https://issues.apache.org/jira/browse/SPARK-28770
>>>>>>>>>>>>>>>>>>>>>>>>>>> [4]
>>>>>>>>>>>>>>>>>>>>>>>>>>> http://status.openlabtesting.org/builds?job_name=spark-master-unit-test-hadoop-2.7-arm64
>>>>>>>>>>>>>>>>>>>>>>>>>>> [5] https://github.com/apache/spark/pull/25186
>>>>>>>>>>>>>>>>>>>>>>>>>>> [6] https://github.com/apache/spark/pull/25279
>>>>>>>>>>>>>>>>>>>>>>>>>>> [7] https://github.com/apache/spark/pull/25673
>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>> On Fri, Aug 16, 2019 at 11:24 PM Sean Owen <
>>>>>>>>>>>>>>>>>>>>>>>>>>> srowen@gmail.com> wrote:
>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>> Yes, I think it's just local caching. After you
>>>>>>>>>>>>>>>>>>>>>>>>>>>> run the build you should find lots of stuff cached at ~/.m2/repository and
>>>>>>>>>>>>>>>>>>>>>>>>>>>> it won't download every time.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>> On Fri, Aug 16, 2019 at 3:01 AM bo zhaobo <
>>>>>>>>>>>>>>>>>>>>>>>>>>>> bzhaojyathousandy@gmail.com> wrote:
>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Hi Sean,
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Thanks for reply. And very apologize for
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> making you confused.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> I know the dependencies will be downloaded
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> from SBT or Maven. But the Spark QA job also exec "mvn clean package", why
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> the log didn't print "downloading some jar from Maven central [1] and build
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> very fast. Is the reason that Spark Jenkins build the Spark jars in the
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> physical machiines and won't destrory the test env after job is finished?
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Then the other job build Spark will get the dependencies jar from the local
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> cached, as the previous jobs exec "mvn package", those dependencies had
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> been downloaded already on local worker machine. Am I right? Is that the
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> reason the job log[1] didn't print any downloading information from Maven
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Central?
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Thank you very much.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> [1]
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/job/spark-master-test-maven-hadoop-2.6-ubuntu-testing/lastBuild/consoleFull
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Best regards
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> ZhaoBo
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> [image: Mailtrack]
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> notified by
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Mailtrack
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/08/16
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> 下午03:58:53
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Sean Owen <sr...@gmail.com> 于2019年8月16日周五
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> 上午10:38写道：
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> I'm not sure what you mean. The dependencies
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> are downloaded by SBT and Maven like in any other project, and nothing
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> about it is specific to Spark.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> The worker machines cache artifacts that are
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> downloaded from these, but this is a function of Maven and SBT, not Spark.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> You may find that the initial download takes a long time.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> On Thu, Aug 15, 2019 at 9:02 PM bo zhaobo <
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> bzhaojyathousandy@gmail.com> wrote:
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Hi Sean,
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Thanks very much for pointing out the
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> roadmap. ;-). Then I think we will continue to focus on our test
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> environment.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> For the networking problems, I mean that we
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> can access Maven Central, and jobs cloud download the required jar package
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> with a high network speed. What we want to know is that, why the Spark QA
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> test jobs[1] log shows the job script/maven build seem don't download the
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> jar packages? Could you tell us the reason about that? Thank you. The
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> reason we raise the "networking problems" is that we found a phenomenon
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> during we test, if we execute "mvn clean package" in a new test
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> environment(As in our test environment, we will destory the test VMs after
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> the job is finish), maven will download the dependency jar packages from
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Maven Central, but in this job "spark-master-test-maven-hadoop" [2], from
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> the log, we didn't found it download any jar packages, what the reason
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> about that?
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Also we build the Spark jar with downloading
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> dependencies from Maven Central, it will cost mostly 1 hour. And we found
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> [2] just cost 10min. But if we run "mvn package" in a VM which already exec
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> "mvn package" before, it just cost 14min, looks very closer with [2]. So we
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> suspect that downloading the Jar packages cost so much time. For the goad
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> of ARM CI, we expect the performance of NEW ARM CI could be closer with
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> existing X86 CI, then users could accept it eaiser.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> [1]
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> [2]
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/job/spark-master-test-maven-hadoop-2.6-ubuntu-testing/lastBuild/consoleFull
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Best regards
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> ZhaoBo
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> [image: Mailtrack]
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> notified by
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Mailtrack
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/08/16
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> 上午09:48:43
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Sean Owen <sr...@gmail.com> 于2019年8月15日周四
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> 下午9:58写道：
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> I think the right goal is to fix the
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> remaining issues first. If we set up CI/CD it will only tell us there are
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> still some test failures. If it's stable, and not hard to add to the
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> existing CI/CD, yes it could be done automatically later. You can continue
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> to test on ARM independently for now.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> It sounds indeed like there are some
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> networking problems in the test system if you're not able to download from
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Maven Central. That rarely takes significant time, and there aren't
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> project-specific mirrors here. You might be able to point at a closer
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> public mirror, depending on where you are.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> On Thu, Aug 15, 2019 at 5:43 AM Tianhua
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> huang <hu...@gmail.com> wrote:
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Hi all,
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> I want to discuss spark ARM CI again, we
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> took some tests on arm instance based on master and the job includes
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://github.com/theopenlab/spark/pull/13 and
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> k8s integration
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://github.com/theopenlab/spark/pull/17/ ,
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> there are several things I want to talk about:
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> First, about the failed tests:
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> 1.we have fixed some problems like
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://github.com/apache/spark/pull/25186
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> and
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://github.com/apache/spark/pull/25279,
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> thanks sean owen and others to help us.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> 2.we tried k8s integration test on
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> arm, and met an error: apk fetch hangs, the tests passed after adding
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> '--network host' option for command `docker build`, see:
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://github.com/theopenlab/spark/pull/17/files#diff-5b731b14068240d63a93c393f6f9b1e8R176
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> , the solution refers to
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://github.com/gliderlabs/docker-alpine/issues/307
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> and I don't know whether it happened once in community CI, or maybe we
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> should submit a pr to pass '--network host' when `docker build`?
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> 3.we found there are two tests failed
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> after the commit
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://github.com/apache/spark/pull/23767
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> :
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> ReplayListenerSuite:
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> - ...
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> - End-to-end replay *** FAILED ***
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> "[driver]" did not equal "[1]"
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> (JsonProtocolSuite.scala:622)
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> - End-to-end replay with
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> compression *** FAILED ***
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> "[driver]" did not equal "[1]"
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> (JsonProtocolSuite.scala:622)
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> we tried to revert the commit and
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> then the tests passed, the patch is too big and so sorry we can't find the
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> reason till now, if you are interesting please try it, and it will be very
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> appreciate if someone can help us to figure it out.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Second, about the test time, we increased
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> the flavor of arm instance to 16U16G, but seems there was no significant
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> improvement, the k8s integration test took about one and a half hours, and
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> the QA test(like spark-master-test-maven-hadoop-2.7 community jenkins job)
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> took about seventeen hours(it is too long :(), we suspect that the reason
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> is the performance and network,
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> we split the jobs based on projects such
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> as sql, core and so on, the time can be decrease to about seven hours, see
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://github.com/theopenlab/spark/pull/19 We
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> found the Spark QA tests like
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/ ,
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> it looks all tests seem never download the jar packages from maven centry
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> repo(such as
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://repo.maven.apache.org/maven2/org/opencypher/okapi-api/0.4.2/okapi-api-0.4.2.jar).
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> So we want to know how the jenkins jobs can do that, is there a internal
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> maven repo launched? maybe we can do the same thing to avoid the network
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> connection cost during downloading the dependent jar packages.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Third, the most important thing, it's
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> about ARM CI of spark, we believe that it is necessary, right? And you can
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> see we really made a lot of efforts, now the basic arm build/test jobs is
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> ok, so we suggest to add arm jobs to community, we can set them to novoting
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> firstly, and improve/rich the jobs step by step. Generally, there are two
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> ways in our mind to integrate the ARM CI for spark:
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> 1) We introduce openlab ARM CI into
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> spark as a custom CI system. We provide human resources and test ARM VMs,
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> also we will focus on the ARM related issues about Spark. We will push the
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> PR into community.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> 2) We donate ARM VM resources into
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> existing amplab Jenkins. We still provide human resources, focus on the ARM
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> related issues about Spark and push the PR into community.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Both options, we will provide human
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> resources to maintain, of course it will be great if we can work together.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> So please tell us which option you would like? And let's move forward.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Waiting for your reply, thank you very much.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>> --
>>>>>>>>>>>>>>>>>>>>>>>> Shane Knapp
>>>>>>>>>>>>>>>>>>>>>>>> UC Berkeley EECS Research / RISELab Staff Technical
>>>>>>>>>>>>>>>>>>>>>>>> Lead
>>>>>>>>>>>>>>>>>>>>>>>> https://rise.cs.berkeley.edu
>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>> --
>>>>>>>>>>>>>>>>>>>>> Shane Knapp
>>>>>>>>>>>>>>>>>>>>> UC Berkeley EECS Research / RISELab Staff Technical
>>>>>>>>>>>>>>>>>>>>> Lead
>>>>>>>>>>>>>>>>>>>>> https://rise.cs.berkeley.edu
>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>> --
>>>>>>>>>>>>>>>>>> Shane Knapp
>>>>>>>>>>>>>>>>>> UC Berkeley EECS Research / RISELab Staff Technical Lead
>>>>>>>>>>>>>>>>>> https://rise.cs.berkeley.edu
>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> --
>>>>>>>>>>>>>>>> Shane Knapp
>>>>>>>>>>>>>>>> UC Berkeley EECS Research / RISELab Staff Technical Lead
>>>>>>>>>>>>>>>> https://rise.cs.berkeley.edu
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> --
>>>>>>>>>>>>>> Shane Knapp
>>>>>>>>>>>>>> UC Berkeley EECS Research / RISELab Staff Technical Lead
>>>>>>>>>>>>>> https://rise.cs.berkeley.edu
>>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>> --
>>>>>>>>>>>>> Shane Knapp
>>>>>>>>>>>>> UC Berkeley EECS Research / RISELab Staff Technical Lead
>>>>>>>>>>>>> https://rise.cs.berkeley.edu
>>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> --
>>>>>>>>>> Shane Knapp
>>>>>>>>>> UC Berkeley EECS Research / RISELab Staff Technical Lead
>>>>>>>>>> https://rise.cs.berkeley.edu
>>>>>>>>>>
>>>>>>>>>
>>>>>>>
>>>>>>> --
>>>>>>> Shane Knapp
>>>>>>> UC Berkeley EECS Research / RISELab Staff Technical Lead
>>>>>>> https://rise.cs.berkeley.edu
>>>>>>>
>>>>>>
>>>>>>
>>>>>> --
>>>>>> Shane Knapp
>>>>>> UC Berkeley EECS Research / RISELab Staff Technical Lead
>>>>>> https://rise.cs.berkeley.edu
>>>>>>
>>>>>
>>>>> [image: Mailtrack]
>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>> notified by
>>>>> Mailtrack
>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/10/16
>>>>> 上午09:13:46
>>>>>
>>>>
>>>>
>>>> --
>>>> Shane Knapp
>>>> UC Berkeley EECS Research / RISELab Staff Technical Lead
>>>> https://rise.cs.berkeley.edu
>>>>
>>>
Re: Ask for ARM CI for spark
Posted by Tianhua huang <hu...@gmail.com>.
Hi all,
Spark arm jobs have built for some time, and now there are two jobs[1]
spark-master-test-maven-arm
<https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-maven-arm/>
and spark-master-test-python-arm
<https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-python-arm/>,
we can see there are some build failures, but it because of the poor
performance of the arm instance, and now we begin to build spark arm jobs
on other high performance instances, and the build/test are all success, we
plan to donate the instance to amplab later. According to the build
history, we are very happy to say spark is supported on aarch64 platform,
and I suggest to add this good news into spark-3.0.0 releasenotes. Maybe
community could provide an arm-supported release of spark at the meanwhile?
[1]
https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-maven-arm/
https://amplab.cs.berkeley.edu/jenkins/job/spark-master-test-python-arm/
ps: the jira https://issues.apache.org/jira/browse/SPARK-29106 trace the
whole work, thank you very much Shane:)
On Thu, Oct 17, 2019 at 2:52 PM bo zhaobo <bz...@gmail.com>
wrote:
> Just Notes: The jira issue link is
> https://issues.apache.org/jira/browse/SPARK-29106
>
>
>
> [image: Mailtrack]
> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
> notified by
> Mailtrack
> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/10/17
> 下午02:50:01
>
> Tianhua huang <hu...@gmail.com> 于2019年10月17日周四 上午10:47写道：
>
>> OK, let's update infos there. Thanks.
>>
>> On Thu, Oct 17, 2019 at 1:52 AM Shane Knapp <sk...@berkeley.edu> wrote:
>>
>>> i totally missed the spark jira from earlier... let's move the
>>> conversation there!
>>>
>>> On Tue, Oct 15, 2019 at 6:21 PM bo zhaobo <bz...@gmail.com>
>>> wrote:
>>>
>>>> Shane, Awaresome! We will try the best to finish the test and the
>>>> requests on the VM recently. Once we finish those things, we will send you
>>>> an email , then we can continue the following things. Thank you very much.
>>>>
>>>> Best Regards,
>>>>
>>>> ZhaoBo
>>>>
>>>> Shane Knapp <sk...@berkeley.edu> 于 2019年10月16日周三 上午3:47写道：
>>>>
>>>>> ok! i'm able to successfully log in to the VM!
>>>>>
>>>>> i also have created a jenkins worker entry:
>>>>> https://amplab.cs.berkeley.edu/jenkins/computer/spark-arm-vm/
>>>>>
>>>>> it's a pretty bare-bones VM, so i have some suggestions/requests
>>>>> before we can actually proceed w/testing. i will not be able to perform
>>>>> any system configuration, as i don't have the cycles to reverse-engineer
>>>>> the ansible setup and test it all out.
>>>>>
>>>>> * java is not installed, please install the following:
>>>>> - java8 min version 1.8.0_191
>>>>> - java11 min version 11.0.1
>>>>>
>>>>> * it appears from the ansible playbook that there are other deps that
>>>>> need to be installed.
>>>>> - please install all deps
>>>>> - manually run the tests until they pass
>>>>>
>>>>> * the jenkins user should NEVER have sudo or any root-level access!
>>>>>
>>>>> * once the arm tests pass when manually run, take a snapshot of this
>>>>> image so we can recreate it w/o needing to reinstall everything
>>>>>
>>>>> after that's done i can finish configuring the jenkins worker and set
>>>>> up a build...
>>>>>
>>>>> thanks!
>>>>>
>>>>> shane
>>>>>
>>>>>
>>>>> On Mon, Oct 14, 2019 at 8:34 PM Shane Knapp <sk...@berkeley.edu>
>>>>> wrote:
>>>>>
>>>>>> yes, i will get to that tomorrow. today was spent cleaning up the
>>>>>> mess from last week.
>>>>>>
>>>>>> On Mon, Oct 14, 2019 at 6:18 PM bo zhaobo <
>>>>>> bzhaojyathousandy@gmail.com> wrote:
>>>>>>
>>>>>>> Hi shane,
>>>>>>>
>>>>>>> That's great news for Amplab is back. ;-) . If possible, could you
>>>>>>> please take several minutes to check the ARM VM is accessible from your
>>>>>>> side? And is there any plan for the whole ARM test integration from
>>>>>>> you?(how about we finish it this month?) Thanks.
>>>>>>>
>>>>>>> Best regards,
>>>>>>>
>>>>>>> ZhaoBo
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> [image: Mailtrack]
>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>>> notified by
>>>>>>> Mailtrack
>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/10/15
>>>>>>> 上午09:13:33
>>>>>>>
>>>>>>> bo zhaobo <bz...@gmail.com> 于2019年10月10日周四 上午8:29写道：
>>>>>>>
>>>>>>>> Oh, sorry about we miss that email. If possible, could you please
>>>>>>>> take some minutes to test the ARM VM is accessible through your ssh private
>>>>>>>> key with jenkins user? And we plan to make the whole integration process
>>>>>>>> and test could be done before the end of this month. We are very happy to
>>>>>>>> work together with you to move it forward, if you are free and agree that.
>>>>>>>> :) Thank you very much.
>>>>>>>>
>>>>>>>> Best Regards,
>>>>>>>> Zhao Bo
>>>>>>>>
>>>>>>>>
>>>>>>>> Shane Knapp <sk...@berkeley.edu> 于 2019年10月9日周三 下午11:10写道：
>>>>>>>>
>>>>>>>>> i spent yesterday dealing w/a power outage on campus. please see
>>>>>>>>> my email to the spark dev list.
>>>>>>>>>
>>>>>>>>> On Wed, Oct 9, 2019 at 3:29 AM Tianhua huang <
>>>>>>>>> huangtianhua223@gmail.com> wrote:
>>>>>>>>>
>>>>>>>>>> Hi Shane,
>>>>>>>>>> Sorry to disturb you again, I wonder if there is a progress of
>>>>>>>>>> the jenkins arm job?
>>>>>>>>>> And please don't hesitate to contact us if you have any questions
>>>>>>>>>> and need helps, thank you very much :)
>>>>>>>>>>
>>>>>>>>>> On Tue, Oct 8, 2019 at 9:21 AM bo zhaobo <
>>>>>>>>>> bzhaojyathousandy@gmail.com> wrote:
>>>>>>>>>>
>>>>>>>>>>> Hi Shane,
>>>>>>>>>>>
>>>>>>>>>>> Sorry for late. We are just back for a 7-days holiday.
>>>>>>>>>>>
>>>>>>>>>>> I already make the public key into the VM. If you are free,
>>>>>>>>>>> please go ahead to test. Thank you
>>>>>>>>>>>
>>>>>>>>>>> Best regards
>>>>>>>>>>>
>>>>>>>>>>> ZhaoBo
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> [image: Mailtrack]
>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>>>>>>> notified by
>>>>>>>>>>> Mailtrack
>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/10/08
>>>>>>>>>>> 上午09:16:49
>>>>>>>>>>>
>>>>>>>>>>> Shane Knapp <sk...@berkeley.edu> 于2019年10月1日周二 上午2:33写道：
>>>>>>>>>>>
>>>>>>>>>>>> here's the public key for the jenkins user:
>>>>>>>>>>>> ssh-rsa
>>>>>>>>>>>> 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
>>>>>>>>>>>> jenkins@amp-jenkins-master
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>> once you get that set up, please let me know and i will test
>>>>>>>>>>>> ssh-ability.
>>>>>>>>>>>>
>>>>>>>>>>>> On Wed, Sep 25, 2019 at 10:15 PM Shane Knapp <
>>>>>>>>>>>> sknapp@berkeley.edu> wrote:
>>>>>>>>>>>>
>>>>>>>>>>>>> thanks for understanding. i'm a one-man show. :)
>>>>>>>>>>>>>
>>>>>>>>>>>>> shane
>>>>>>>>>>>>>
>>>>>>>>>>>>> On Wed, Sep 25, 2019 at 6:52 PM Tianhua huang <
>>>>>>>>>>>>> huangtianhua223@gmail.com> wrote:
>>>>>>>>>>>>>
>>>>>>>>>>>>>> I am very happy to hear that :)
>>>>>>>>>>>>>> And please don't hesitate to contact us if you have any
>>>>>>>>>>>>>> questions and need helps, thanks again.
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> On Thu, Sep 26, 2019 at 12:34 AM Shane Knapp <
>>>>>>>>>>>>>> sknapp@berkeley.edu> wrote:
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> i'm hoping to get started by EOW or beginning of next. i
>>>>>>>>>>>>>>> have an incredibly small team here @ berkeley and support a lot of research
>>>>>>>>>>>>>>> labs, and i've been down 25-50% staffing for the past 6 weeks. thankfully
>>>>>>>>>>>>>>> my whole team is back and i'm finally getting my head above water and will
>>>>>>>>>>>>>>> have time to dedicate to this really soon.
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> sorry for the delay!
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> shane
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> On Mon, Sep 23, 2019 at 7:21 PM bo zhaobo <
>>>>>>>>>>>>>>> bzhaojyathousandy@gmail.com> wrote:
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> Hi Shane,
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> How about to start the ARM work this week? Can we? ;-).
>>>>>>>>>>>>>>>> Thank you
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> Best Regards
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> [image: Mailtrack]
>>>>>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>>>>>>>>>>>> notified by
>>>>>>>>>>>>>>>> Mailtrack
>>>>>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/09/24
>>>>>>>>>>>>>>>> 上午10:18:01
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> Shane Knapp <sk...@berkeley.edu> 于2019年9月20日周五 上午11:57写道：
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>> i'll have the cycles over the next week.
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>> On Thu, Sep 19, 2019 at 7:51 PM bo zhaobo <
>>>>>>>>>>>>>>>>> bzhaojyathousandy@gmail.com> wrote:
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>> Hi Shane,
>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>> Is there any update about the last email? ;-)
>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>> Thanks,
>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>> Best Regards
>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>> [image: Mailtrack]
>>>>>>>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>>>>>>>>>>>>>> notified by
>>>>>>>>>>>>>>>>>> Mailtrack
>>>>>>>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/09/20
>>>>>>>>>>>>>>>>>> 上午10:49:35
>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>> bo zhaobo <bz...@gmail.com> 于2019年9月18日周三
>>>>>>>>>>>>>>>>>> 上午10:02写道：
>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>> Thanks for reply.
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>> I will answer you question One by One.
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>> 0) Sure, I can create an user named 'jenkins' manually.
>>>>>>>>>>>>>>>>>>> And I will left the 'arm' User if you might need the higher priority to do
>>>>>>>>>>>>>>>>>>> something on the VM.
>>>>>>>>>>>>>>>>>>> 1) Sure, that would be great if you could provide a
>>>>>>>>>>>>>>>>>>> public key to us. ;-)
>>>>>>>>>>>>>>>>>>> 2) Sure
>>>>>>>>>>>>>>>>>>> 3) Yeah, it's a persistent VM now. But from us, we plan
>>>>>>>>>>>>>>>>>>> to donate more resources beside the end of October. It's better to move the
>>>>>>>>>>>>>>>>>>> all ARM jenkins workers to the same place, so this VM might be recycle in
>>>>>>>>>>>>>>>>>>> the future. For now, that dosen't stop us to move the CI job forward.
>>>>>>>>>>>>>>>>>>> 4) Correct. ;-)
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>> If you need more information or help about the ARM VM,
>>>>>>>>>>>>>>>>>>> please free to contact us.
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>> Questions from us:
>>>>>>>>>>>>>>>>>>> 1> is there a online chat platform? then we can chat and
>>>>>>>>>>>>>>>>>>> discuss on that. Seem emailing is too slow for us. ;-). Or is there a
>>>>>>>>>>>>>>>>>>> online discuss platform in Spark community?
>>>>>>>>>>>>>>>>>>> 2> Could you please share the test scripts once you
>>>>>>>>>>>>>>>>>>> finish the test integration? Also the existing Jenkins test scripts in
>>>>>>>>>>>>>>>>>>> Spark CI, as we plan to do the same thing like X86, so We think testing the
>>>>>>>>>>>>>>>>>>> same jobs like X86 on ARM would be a good start.
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>> Thank you
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>> Best Regards
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>> [image: Mailtrack]
>>>>>>>>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>>>>>>>>>>>>>>> notified by
>>>>>>>>>>>>>>>>>>> Mailtrack
>>>>>>>>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/09/18
>>>>>>>>>>>>>>>>>>> 上午09:33:53
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>> Shane Knapp <sk...@berkeley.edu> 于2019年9月18日周三
>>>>>>>>>>>>>>>>>>> 上午5:05写道：
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>> thanks for the info... a couple of things:
>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>> 0) could you rename/create a 'jenkins' user with the
>>>>>>>>>>>>>>>>>>>> same rights?
>>>>>>>>>>>>>>>>>>>> 1) i will provide an ssh key so that you can add it to
>>>>>>>>>>>>>>>>>>>> the jenkins user's authorized_keys file
>>>>>>>>>>>>>>>>>>>> 2) the jenkins user should NOT have root access. this
>>>>>>>>>>>>>>>>>>>> is a major security hole
>>>>>>>>>>>>>>>>>>>> 3) will this be a persistent VM? if so, i'd much
>>>>>>>>>>>>>>>>>>>> prefer to have it set up initially so we can just log in, build what we
>>>>>>>>>>>>>>>>>>>> need and launch the job
>>>>>>>>>>>>>>>>>>>> 4) yay ansible! i can create the job locally from the
>>>>>>>>>>>>>>>>>>>> template.
>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>> On Tue, Sep 17, 2019 at 1:16 AM bo zhaobo <
>>>>>>>>>>>>>>>>>>>> bzhaojyathousandy@gmail.com> wrote:
>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>> Hi @shane knapp <sk...@berkeley.edu> , Thanks.
>>>>>>>>>>>>>>>>>>>>> Tianhua huang and I already create a ARM VM for this.
>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>> The Infomation about the ARM VM:
>>>>>>>>>>>>>>>>>>>>> IP: 213.146.141.16
>>>>>>>>>>>>>>>>>>>>> ssh_key: please see the attachment file
>>>>>>>>>>>>>>>>>>>>> "lab_ssh_key.txt"
>>>>>>>>>>>>>>>>>>>>> Then you can login to it with "ssh -i lab_ssh_key.txt
>>>>>>>>>>>>>>>>>>>>> arm@213.146.141.16 ", the User "arm" have sudo
>>>>>>>>>>>>>>>>>>>>> priority.
>>>>>>>>>>>>>>>>>>>>> Notes: This VM is running on Cloud, now it just enable
>>>>>>>>>>>>>>>>>>>>> ssh (tcp 22 port) and ping (icmp) for ingress network connection, if you
>>>>>>>>>>>>>>>>>>>>> need more, please let us know.
>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>> And first we can add periodic job for spark arm test,
>>>>>>>>>>>>>>>>>>>>> maybe the zuul job details of openlab we tested can be referenced in
>>>>>>>>>>>>>>>>>>>>> jenkins job
>>>>>>>>>>>>>>>>>>>>> https://github.com/theopenlab/openlab-zuul-jobs/blob/master/playbooks/spark-unit-test-arm64/run.yaml
>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>> Let's give you some explaination about that file. In
>>>>>>>>>>>>>>>>>>>>> the Step3 of above file, leveldbjni doesn't have a ARM release, so we build
>>>>>>>>>>>>>>>>>>>>> leveldbjni on arm and use ourselves leveldbjni jar, which locates on [1].
>>>>>>>>>>>>>>>>>>>>> So you need to still use the jar to test. And hadoop side also depends on
>>>>>>>>>>>>>>>>>>>>> leveldbjni, I remember that is "hadoop-client". For this reason, we can not
>>>>>>>>>>>>>>>>>>>>> change the pom file directly, so we use "mvn install" to enable leveldb
>>>>>>>>>>>>>>>>>>>>> during testing spark on ARM.
>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>> Please Notes： All test runs without ROOT user. So
>>>>>>>>>>>>>>>>>>>>> that's good if you test with "arm" User.
>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>> Thanks,
>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>> Best Regards
>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>> [1]
>>>>>>>>>>>>>>>>>>>>> https://repo1.maven.org/maven2/org/openlabtesting/leveldbjni/leveldbjni-all/1.8/leveldbjni-all-1.8.jar
>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>> [image: Mailtrack]
>>>>>>>>>>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>>>>>>>>>>>>>>>>> notified by
>>>>>>>>>>>>>>>>>>>>> Mailtrack
>>>>>>>>>>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/09/17
>>>>>>>>>>>>>>>>>>>>> 下午04:08:59
>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>> Tianhua huang <hu...@gmail.com>
>>>>>>>>>>>>>>>>>>>>> 于2019年9月17日周二 下午2:48写道：
>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>> @shane knapp <sk...@berkeley.edu> thank you very
>>>>>>>>>>>>>>>>>>>>>> much, I opened an issue for this
>>>>>>>>>>>>>>>>>>>>>> https://issues.apache.org/jira/browse/SPARK-29106,
>>>>>>>>>>>>>>>>>>>>>> we can tall the details in it :)
>>>>>>>>>>>>>>>>>>>>>> And we will prepare an arm instance today and will
>>>>>>>>>>>>>>>>>>>>>> send the info to your email later.
>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>> On Tue, Sep 17, 2019 at 4:40 AM Shane Knapp <
>>>>>>>>>>>>>>>>>>>>>> sknapp@berkeley.edu> wrote:
>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>> @Tianhua huang <hu...@gmail.com> sure, i
>>>>>>>>>>>>>>>>>>>>>>> think we can get something sorted for the short-term.
>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>> all we need is ssh access (i can provide an ssh
>>>>>>>>>>>>>>>>>>>>>>> key), and i can then have our jenkins master launch a remote worker on that
>>>>>>>>>>>>>>>>>>>>>>> instance.
>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>> instance setup, etc, will be up to you. my support
>>>>>>>>>>>>>>>>>>>>>>> for the time being will be to create the job and 'best effort' for
>>>>>>>>>>>>>>>>>>>>>>> everything else.
>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>> this should get us up and running asap.
>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>> is there an open JIRA for jenkins/arm test support?
>>>>>>>>>>>>>>>>>>>>>>> we can move the technical details about this idea there.
>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>> On Sun, Sep 15, 2019 at 9:03 PM Tianhua huang <
>>>>>>>>>>>>>>>>>>>>>>> huangtianhua223@gmail.com> wrote:
>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>> @Sean Owen <sr...@gmail.com> , so sorry to reply
>>>>>>>>>>>>>>>>>>>>>>>> late, we had a Mid-Autumn holiday:)
>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>> If you hope to integrate ARM CI to amplab jenkins,
>>>>>>>>>>>>>>>>>>>>>>>> we can offer the arm instance, and then the ARM job will run together with
>>>>>>>>>>>>>>>>>>>>>>>> other x86 jobs, so maybe there is a guideline to do this? @shane
>>>>>>>>>>>>>>>>>>>>>>>> knapp <sk...@berkeley.edu> would you help us?
>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>> On Thu, Sep 12, 2019 at 9:36 PM Sean Owen <
>>>>>>>>>>>>>>>>>>>>>>>> srowen@gmail.com> wrote:
>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>> I don't know what's involved in actually accepting
>>>>>>>>>>>>>>>>>>>>>>>>> or operating those machines, so can't comment there, but in the meantime
>>>>>>>>>>>>>>>>>>>>>>>>> it's good that you are running these tests and can help report changes
>>>>>>>>>>>>>>>>>>>>>>>>> needed to keep it working with ARM. I would continue with that for now.
>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>> On Wed, Sep 11, 2019 at 10:06 PM Tianhua huang <
>>>>>>>>>>>>>>>>>>>>>>>>> huangtianhua223@gmail.com> wrote:
>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>> Hi all,
>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>> For the whole work process of spark ARM CI, we
>>>>>>>>>>>>>>>>>>>>>>>>>> want to make 2 things clear.
>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>> The first thing is:
>>>>>>>>>>>>>>>>>>>>>>>>>> About spark ARM CI, now we have two periodic
>>>>>>>>>>>>>>>>>>>>>>>>>> jobs, one job[1] based on commit[2](which already fixed the replay tests
>>>>>>>>>>>>>>>>>>>>>>>>>> failed issue[3], we made a new test branch based on date 09-09-2019), the
>>>>>>>>>>>>>>>>>>>>>>>>>> other job[4] based on spark master.
>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>> The first job we test on the specified branch to
>>>>>>>>>>>>>>>>>>>>>>>>>> prove that our ARM CI is good and stable.
>>>>>>>>>>>>>>>>>>>>>>>>>> The second job checks spark master every day,
>>>>>>>>>>>>>>>>>>>>>>>>>> then we can find whether the latest commits affect the ARM CI. According to
>>>>>>>>>>>>>>>>>>>>>>>>>> the build history and result, it shows that some problems are easier to
>>>>>>>>>>>>>>>>>>>>>>>>>> find on ARM like SPARK-28770
>>>>>>>>>>>>>>>>>>>>>>>>>> <https://issues.apache.org/jira/browse/SPARK-28770>,
>>>>>>>>>>>>>>>>>>>>>>>>>> and it also shows that we would make efforts to trace and figure them
>>>>>>>>>>>>>>>>>>>>>>>>>> out, till now we have found and fixed several problems[5][6][7], thanks
>>>>>>>>>>>>>>>>>>>>>>>>>> everyone of the community :). And we believe that ARM CI is very necessary,
>>>>>>>>>>>>>>>>>>>>>>>>>> right?
>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>> The second thing is:
>>>>>>>>>>>>>>>>>>>>>>>>>> We plan to run the jobs for a period of time, and
>>>>>>>>>>>>>>>>>>>>>>>>>> you can see the result and logs from 'build history' of the jobs console,
>>>>>>>>>>>>>>>>>>>>>>>>>> if everything goes well for one or two weeks could community accept the ARM
>>>>>>>>>>>>>>>>>>>>>>>>>> CI? or how long the periodic jobs to run then our community could have
>>>>>>>>>>>>>>>>>>>>>>>>>> enough confidence to accept the ARM CI? As you suggested before, it's good
>>>>>>>>>>>>>>>>>>>>>>>>>> to integrate ARM CI to amplab jenkins, we agree that and we can donate the
>>>>>>>>>>>>>>>>>>>>>>>>>> ARM instances and then maintain the ARM-related test jobs together with
>>>>>>>>>>>>>>>>>>>>>>>>>> community, any thoughts?
>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>> Thank you all!
>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>> [1]
>>>>>>>>>>>>>>>>>>>>>>>>>> http://status.openlabtesting.org/job/spark-unchanged-branch-unit-test-hadoop-2.7-arm64
>>>>>>>>>>>>>>>>>>>>>>>>>> [2]
>>>>>>>>>>>>>>>>>>>>>>>>>> https://github.com/apache/spark/commit/0ed9fae45769d4b06b8cf8128f462f09ff3d9a72
>>>>>>>>>>>>>>>>>>>>>>>>>> [3]
>>>>>>>>>>>>>>>>>>>>>>>>>> https://issues.apache.org/jira/browse/SPARK-28770
>>>>>>>>>>>>>>>>>>>>>>>>>> [4]
>>>>>>>>>>>>>>>>>>>>>>>>>> http://status.openlabtesting.org/builds?job_name=spark-master-unit-test-hadoop-2.7-arm64
>>>>>>>>>>>>>>>>>>>>>>>>>> [5] https://github.com/apache/spark/pull/25186
>>>>>>>>>>>>>>>>>>>>>>>>>> [6] https://github.com/apache/spark/pull/25279
>>>>>>>>>>>>>>>>>>>>>>>>>> [7] https://github.com/apache/spark/pull/25673
>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>> On Fri, Aug 16, 2019 at 11:24 PM Sean Owen <
>>>>>>>>>>>>>>>>>>>>>>>>>> srowen@gmail.com> wrote:
>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>> Yes, I think it's just local caching. After you
>>>>>>>>>>>>>>>>>>>>>>>>>>> run the build you should find lots of stuff cached at ~/.m2/repository and
>>>>>>>>>>>>>>>>>>>>>>>>>>> it won't download every time.
>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>> On Fri, Aug 16, 2019 at 3:01 AM bo zhaobo <
>>>>>>>>>>>>>>>>>>>>>>>>>>> bzhaojyathousandy@gmail.com> wrote:
>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>> Hi Sean,
>>>>>>>>>>>>>>>>>>>>>>>>>>>> Thanks for reply. And very apologize for making
>>>>>>>>>>>>>>>>>>>>>>>>>>>> you confused.
>>>>>>>>>>>>>>>>>>>>>>>>>>>> I know the dependencies will be downloaded from
>>>>>>>>>>>>>>>>>>>>>>>>>>>> SBT or Maven. But the Spark QA job also exec "mvn clean package", why the
>>>>>>>>>>>>>>>>>>>>>>>>>>>> log didn't print "downloading some jar from Maven central [1] and build
>>>>>>>>>>>>>>>>>>>>>>>>>>>> very fast. Is the reason that Spark Jenkins build the Spark jars in the
>>>>>>>>>>>>>>>>>>>>>>>>>>>> physical machiines and won't destrory the test env after job is finished?
>>>>>>>>>>>>>>>>>>>>>>>>>>>> Then the other job build Spark will get the dependencies jar from the local
>>>>>>>>>>>>>>>>>>>>>>>>>>>> cached, as the previous jobs exec "mvn package", those dependencies had
>>>>>>>>>>>>>>>>>>>>>>>>>>>> been downloaded already on local worker machine. Am I right? Is that the
>>>>>>>>>>>>>>>>>>>>>>>>>>>> reason the job log[1] didn't print any downloading information from Maven
>>>>>>>>>>>>>>>>>>>>>>>>>>>> Central?
>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>> Thank you very much.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>> [1]
>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/job/spark-master-test-maven-hadoop-2.6-ubuntu-testing/lastBuild/consoleFull
>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>> Best regards
>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>> ZhaoBo
>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>> [image: Mailtrack]
>>>>>>>>>>>>>>>>>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>>>>>>>>>>>>>>>>>>>>>>>> notified by
>>>>>>>>>>>>>>>>>>>>>>>>>>>> Mailtrack
>>>>>>>>>>>>>>>>>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/08/16
>>>>>>>>>>>>>>>>>>>>>>>>>>>> 下午03:58:53
>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>> Sean Owen <sr...@gmail.com> 于2019年8月16日周五
>>>>>>>>>>>>>>>>>>>>>>>>>>>> 上午10:38写道：
>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> I'm not sure what you mean. The dependencies
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> are downloaded by SBT and Maven like in any other project, and nothing
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> about it is specific to Spark.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> The worker machines cache artifacts that are
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> downloaded from these, but this is a function of Maven and SBT, not Spark.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> You may find that the initial download takes a long time.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> On Thu, Aug 15, 2019 at 9:02 PM bo zhaobo <
>>>>>>>>>>>>>>>>>>>>>>>>>>>>> bzhaojyathousandy@gmail.com> wrote:
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Hi Sean,
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Thanks very much for pointing out the
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> roadmap. ;-). Then I think we will continue to focus on our test
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> environment.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> For the networking problems, I mean that we
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> can access Maven Central, and jobs cloud download the required jar package
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> with a high network speed. What we want to know is that, why the Spark QA
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> test jobs[1] log shows the job script/maven build seem don't download the
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> jar packages? Could you tell us the reason about that? Thank you. The
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> reason we raise the "networking problems" is that we found a phenomenon
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> during we test, if we execute "mvn clean package" in a new test
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> environment(As in our test environment, we will destory the test VMs after
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> the job is finish), maven will download the dependency jar packages from
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Maven Central, but in this job "spark-master-test-maven-hadoop" [2], from
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> the log, we didn't found it download any jar packages, what the reason
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> about that?
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Also we build the Spark jar with downloading
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> dependencies from Maven Central, it will cost mostly 1 hour. And we found
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> [2] just cost 10min. But if we run "mvn package" in a VM which already exec
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> "mvn package" before, it just cost 14min, looks very closer with [2]. So we
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> suspect that downloading the Jar packages cost so much time. For the goad
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> of ARM CI, we expect the performance of NEW ARM CI could be closer with
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> existing X86 CI, then users could accept it eaiser.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> [1]
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> [2]
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/job/spark-master-test-maven-hadoop-2.6-ubuntu-testing/lastBuild/consoleFull
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Best regards
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> ZhaoBo
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> [image: Mailtrack]
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> notified by
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Mailtrack
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/08/16
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> 上午09:48:43
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Sean Owen <sr...@gmail.com> 于2019年8月15日周四
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> 下午9:58写道：
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> I think the right goal is to fix the
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> remaining issues first. If we set up CI/CD it will only tell us there are
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> still some test failures. If it's stable, and not hard to add to the
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> existing CI/CD, yes it could be done automatically later. You can continue
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> to test on ARM independently for now.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> It sounds indeed like there are some
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> networking problems in the test system if you're not able to download from
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Maven Central. That rarely takes significant time, and there aren't
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> project-specific mirrors here. You might be able to point at a closer
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> public mirror, depending on where you are.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> On Thu, Aug 15, 2019 at 5:43 AM Tianhua
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> huang <hu...@gmail.com> wrote:
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Hi all,
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> I want to discuss spark ARM CI again, we
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> took some tests on arm instance based on master and the job includes
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://github.com/theopenlab/spark/pull/13 and
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> k8s integration
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://github.com/theopenlab/spark/pull/17/ ,
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> there are several things I want to talk about:
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> First, about the failed tests:
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> 1.we have fixed some problems like
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://github.com/apache/spark/pull/25186
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> and
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://github.com/apache/spark/pull/25279,
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> thanks sean owen and others to help us.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> 2.we tried k8s integration test on arm,
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> and met an error: apk fetch hangs, the tests passed after adding
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> '--network host' option for command `docker build`, see:
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://github.com/theopenlab/spark/pull/17/files#diff-5b731b14068240d63a93c393f6f9b1e8R176
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> , the solution refers to
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://github.com/gliderlabs/docker-alpine/issues/307
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> and I don't know whether it happened once in community CI, or maybe we
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> should submit a pr to pass '--network host' when `docker build`?
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> 3.we found there are two tests failed
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> after the commit
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://github.com/apache/spark/pull/23767
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> :
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> ReplayListenerSuite:
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> - ...
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> - End-to-end replay *** FAILED ***
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> "[driver]" did not equal "[1]"
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> (JsonProtocolSuite.scala:622)
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> - End-to-end replay with compression
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> *** FAILED ***
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> "[driver]" did not equal "[1]"
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> (JsonProtocolSuite.scala:622)
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> we tried to revert the commit and
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> then the tests passed, the patch is too big and so sorry we can't find the
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> reason till now, if you are interesting please try it, and it will be very
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> appreciate if someone can help us to figure it out.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Second, about the test time, we increased
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> the flavor of arm instance to 16U16G, but seems there was no significant
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> improvement, the k8s integration test took about one and a half hours, and
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> the QA test(like spark-master-test-maven-hadoop-2.7 community jenkins job)
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> took about seventeen hours(it is too long :(), we suspect that the reason
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> is the performance and network,
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> we split the jobs based on projects such as
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> sql, core and so on, the time can be decrease to about seven hours, see
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://github.com/theopenlab/spark/pull/19 We
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> found the Spark QA tests like
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/ ,
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> it looks all tests seem never download the jar packages from maven centry
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> repo(such as
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> https://repo.maven.apache.org/maven2/org/opencypher/okapi-api/0.4.2/okapi-api-0.4.2.jar).
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> So we want to know how the jenkins jobs can do that, is there a internal
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> maven repo launched? maybe we can do the same thing to avoid the network
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> connection cost during downloading the dependent jar packages.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Third, the most important thing, it's about
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> ARM CI of spark, we believe that it is necessary, right? And you can see we
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> really made a lot of efforts, now the basic arm build/test jobs is ok, so
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> we suggest to add arm jobs to community, we can set them to novoting
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> firstly, and improve/rich the jobs step by step. Generally, there are two
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> ways in our mind to integrate the ARM CI for spark:
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> 1) We introduce openlab ARM CI into
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> spark as a custom CI system. We provide human resources and test ARM VMs,
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> also we will focus on the ARM related issues about Spark. We will push the
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> PR into community.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> 2) We donate ARM VM resources into
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> existing amplab Jenkins. We still provide human resources, focus on the ARM
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> related issues about Spark and push the PR into community.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Both options, we will provide human
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> resources to maintain, of course it will be great if we can work together.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> So please tell us which option you would like? And let's move forward.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Waiting for your reply, thank you very much.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>> --
>>>>>>>>>>>>>>>>>>>>>>> Shane Knapp
>>>>>>>>>>>>>>>>>>>>>>> UC Berkeley EECS Research / RISELab Staff Technical
>>>>>>>>>>>>>>>>>>>>>>> Lead
>>>>>>>>>>>>>>>>>>>>>>> https://rise.cs.berkeley.edu
>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>> --
>>>>>>>>>>>>>>>>>>>> Shane Knapp
>>>>>>>>>>>>>>>>>>>> UC Berkeley EECS Research / RISELab Staff Technical Lead
>>>>>>>>>>>>>>>>>>>> https://rise.cs.berkeley.edu
>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>> --
>>>>>>>>>>>>>>>>> Shane Knapp
>>>>>>>>>>>>>>>>> UC Berkeley EECS Research / RISELab Staff Technical Lead
>>>>>>>>>>>>>>>>> https://rise.cs.berkeley.edu
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> --
>>>>>>>>>>>>>>> Shane Knapp
>>>>>>>>>>>>>>> UC Berkeley EECS Research / RISELab Staff Technical Lead
>>>>>>>>>>>>>>> https://rise.cs.berkeley.edu
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>> --
>>>>>>>>>>>>> Shane Knapp
>>>>>>>>>>>>> UC Berkeley EECS Research / RISELab Staff Technical Lead
>>>>>>>>>>>>> https://rise.cs.berkeley.edu
>>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>> --
>>>>>>>>>>>> Shane Knapp
>>>>>>>>>>>> UC Berkeley EECS Research / RISELab Staff Technical Lead
>>>>>>>>>>>> https://rise.cs.berkeley.edu
>>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>
>>>>>>>>> --
>>>>>>>>> Shane Knapp
>>>>>>>>> UC Berkeley EECS Research / RISELab Staff Technical Lead
>>>>>>>>> https://rise.cs.berkeley.edu
>>>>>>>>>
>>>>>>>>
>>>>>>
>>>>>> --
>>>>>> Shane Knapp
>>>>>> UC Berkeley EECS Research / RISELab Staff Technical Lead
>>>>>> https://rise.cs.berkeley.edu
>>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> Shane Knapp
>>>>> UC Berkeley EECS Research / RISELab Staff Technical Lead
>>>>> https://rise.cs.berkeley.edu
>>>>>
>>>>
>>>> [image: Mailtrack]
>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>> notified by
>>>> Mailtrack
>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/10/16
>>>> 上午09:13:46
>>>>
>>>
>>>
>>> --
>>> Shane Knapp
>>> UC Berkeley EECS Research / RISELab Staff Technical Lead
>>> https://rise.cs.berkeley.edu
>>>
>>
Re: Ask for ARM CI for spark
Posted by Tianhua huang <hu...@gmail.com>.
@shane knapp <sk...@berkeley.edu> thank you very much, I opened an issue
for this https://issues.apache.org/jira/browse/SPARK-29106, we can tall the
details in it :)
And we will prepare an arm instance today and will send the info to your
email later.
On Tue, Sep 17, 2019 at 4:40 AM Shane Knapp <sk...@berkeley.edu> wrote:
> @Tianhua huang <hu...@gmail.com> sure, i think we can get
> something sorted for the short-term.
>
> all we need is ssh access (i can provide an ssh key), and i can then have
> our jenkins master launch a remote worker on that instance.
>
> instance setup, etc, will be up to you. my support for the time being
> will be to create the job and 'best effort' for everything else.
>
> this should get us up and running asap.
>
> is there an open JIRA for jenkins/arm test support? we can move the
> technical details about this idea there.
>
> On Sun, Sep 15, 2019 at 9:03 PM Tianhua huang <hu...@gmail.com>
> wrote:
>
>> @Sean Owen <sr...@gmail.com> , so sorry to reply late, we had a
>> Mid-Autumn holiday:)
>>
>> If you hope to integrate ARM CI to amplab jenkins, we can offer the arm
>> instance, and then the ARM job will run together with other x86 jobs, so
>> maybe there is a guideline to do this? @shane knapp <sk...@berkeley.edu>
>> would you help us?
>>
>> On Thu, Sep 12, 2019 at 9:36 PM Sean Owen <sr...@gmail.com> wrote:
>>
>>> I don't know what's involved in actually accepting or operating those
>>> machines, so can't comment there, but in the meantime it's good that you
>>> are running these tests and can help report changes needed to keep it
>>> working with ARM. I would continue with that for now.
>>>
>>> On Wed, Sep 11, 2019 at 10:06 PM Tianhua huang <
>>> huangtianhua223@gmail.com> wrote:
>>>
>>>> Hi all,
>>>>
>>>> For the whole work process of spark ARM CI, we want to make 2 things
>>>> clear.
>>>>
>>>> The first thing is:
>>>> About spark ARM CI, now we have two periodic jobs, one job[1] based on
>>>> commit[2](which already fixed the replay tests failed issue[3], we made a
>>>> new test branch based on date 09-09-2019), the other job[4] based on spark
>>>> master.
>>>>
>>>> The first job we test on the specified branch to prove that our ARM CI
>>>> is good and stable.
>>>> The second job checks spark master every day, then we can find whether
>>>> the latest commits affect the ARM CI. According to the build history and
>>>> result, it shows that some problems are easier to find on ARM like
>>>> SPARK-28770 <https://issues.apache.org/jira/browse/SPARK-28770>, and
>>>> it also shows that we would make efforts to trace and figure them out, till
>>>> now we have found and fixed several problems[5][6][7], thanks everyone of
>>>> the community :). And we believe that ARM CI is very necessary, right?
>>>>
>>>> The second thing is:
>>>> We plan to run the jobs for a period of time, and you can see the
>>>> result and logs from 'build history' of the jobs console, if everything
>>>> goes well for one or two weeks could community accept the ARM CI? or how
>>>> long the periodic jobs to run then our community could have enough
>>>> confidence to accept the ARM CI? As you suggested before, it's good to
>>>> integrate ARM CI to amplab jenkins, we agree that and we can donate the ARM
>>>> instances and then maintain the ARM-related test jobs together with
>>>> community, any thoughts?
>>>>
>>>> Thank you all!
>>>>
>>>> [1]
>>>> http://status.openlabtesting.org/job/spark-unchanged-branch-unit-test-hadoop-2.7-arm64
>>>> [2]
>>>> https://github.com/apache/spark/commit/0ed9fae45769d4b06b8cf8128f462f09ff3d9a72
>>>> [3] https://issues.apache.org/jira/browse/SPARK-28770
>>>> [4]
>>>> http://status.openlabtesting.org/builds?job_name=spark-master-unit-test-hadoop-2.7-arm64
>>>> [5] https://github.com/apache/spark/pull/25186
>>>> [6] https://github.com/apache/spark/pull/25279
>>>> [7] https://github.com/apache/spark/pull/25673
>>>>
>>>>
>>>>
>>>> On Fri, Aug 16, 2019 at 11:24 PM Sean Owen <sr...@gmail.com> wrote:
>>>>
>>>>> Yes, I think it's just local caching. After you run the build you
>>>>> should find lots of stuff cached at ~/.m2/repository and it won't download
>>>>> every time.
>>>>>
>>>>> On Fri, Aug 16, 2019 at 3:01 AM bo zhaobo <bz...@gmail.com>
>>>>> wrote:
>>>>>
>>>>>> Hi Sean,
>>>>>> Thanks for reply. And very apologize for making you confused.
>>>>>> I know the dependencies will be downloaded from SBT or Maven. But the
>>>>>> Spark QA job also exec "mvn clean package", why the log didn't print
>>>>>> "downloading some jar from Maven central [1] and build very fast. Is the
>>>>>> reason that Spark Jenkins build the Spark jars in the physical machiines
>>>>>> and won't destrory the test env after job is finished? Then the other job
>>>>>> build Spark will get the dependencies jar from the local cached, as the
>>>>>> previous jobs exec "mvn package", those dependencies had been downloaded
>>>>>> already on local worker machine. Am I right? Is that the reason the job
>>>>>> log[1] didn't print any downloading information from Maven Central?
>>>>>>
>>>>>> Thank you very much.
>>>>>>
>>>>>> [1]
>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/job/spark-master-test-maven-hadoop-2.6-ubuntu-testing/lastBuild/consoleFull
>>>>>>
>>>>>>
>>>>>> Best regards
>>>>>>
>>>>>> ZhaoBo
>>>>>>
>>>>>> [image: Mailtrack]
>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>> notified by
>>>>>> Mailtrack
>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/08/16
>>>>>> 下午03:58:53
>>>>>>
>>>>>> Sean Owen <sr...@gmail.com> 于2019年8月16日周五 上午10:38写道：
>>>>>>
>>>>>>> I'm not sure what you mean. The dependencies are downloaded by SBT
>>>>>>> and Maven like in any other project, and nothing about it is specific to
>>>>>>> Spark.
>>>>>>> The worker machines cache artifacts that are downloaded from these,
>>>>>>> but this is a function of Maven and SBT, not Spark. You may find that the
>>>>>>> initial download takes a long time.
>>>>>>>
>>>>>>> On Thu, Aug 15, 2019 at 9:02 PM bo zhaobo <
>>>>>>> bzhaojyathousandy@gmail.com> wrote:
>>>>>>>
>>>>>>>> Hi Sean,
>>>>>>>>
>>>>>>>> Thanks very much for pointing out the roadmap. ;-). Then I think we
>>>>>>>> will continue to focus on our test environment.
>>>>>>>>
>>>>>>>> For the networking problems, I mean that we can access Maven
>>>>>>>> Central, and jobs cloud download the required jar package with a high
>>>>>>>> network speed. What we want to know is that, why the Spark QA test jobs[1]
>>>>>>>> log shows the job script/maven build seem don't download the jar packages?
>>>>>>>> Could you tell us the reason about that? Thank you. The reason we raise
>>>>>>>> the "networking problems" is that we found a phenomenon during we test, if
>>>>>>>> we execute "mvn clean package" in a new test environment(As in our test
>>>>>>>> environment, we will destory the test VMs after the job is finish), maven
>>>>>>>> will download the dependency jar packages from Maven Central, but in this
>>>>>>>> job "spark-master-test-maven-hadoop" [2], from the log, we didn't found it
>>>>>>>> download any jar packages, what the reason about that?
>>>>>>>> Also we build the Spark jar with downloading dependencies from
>>>>>>>> Maven Central, it will cost mostly 1 hour. And we found [2] just cost
>>>>>>>> 10min. But if we run "mvn package" in a VM which already exec "mvn package"
>>>>>>>> before, it just cost 14min, looks very closer with [2]. So we suspect that
>>>>>>>> downloading the Jar packages cost so much time. For the goad of ARM CI, we
>>>>>>>> expect the performance of NEW ARM CI could be closer with existing X86 CI,
>>>>>>>> then users could accept it eaiser.
>>>>>>>>
>>>>>>>> [1] https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/
>>>>>>>> [2]
>>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/job/spark-master-test-maven-hadoop-2.6-ubuntu-testing/lastBuild/consoleFull
>>>>>>>>
>>>>>>>> Best regards
>>>>>>>>
>>>>>>>> ZhaoBo
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> [image: Mailtrack]
>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>>>> notified by
>>>>>>>> Mailtrack
>>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/08/16
>>>>>>>> 上午09:48:43
>>>>>>>>
>>>>>>>> Sean Owen <sr...@gmail.com> 于2019年8月15日周四 下午9:58写道：
>>>>>>>>
>>>>>>>>> I think the right goal is to fix the remaining issues first. If we
>>>>>>>>> set up CI/CD it will only tell us there are still some test failures. If
>>>>>>>>> it's stable, and not hard to add to the existing CI/CD, yes it could be
>>>>>>>>> done automatically later. You can continue to test on ARM independently for
>>>>>>>>> now.
>>>>>>>>>
>>>>>>>>> It sounds indeed like there are some networking problems in the
>>>>>>>>> test system if you're not able to download from Maven Central. That rarely
>>>>>>>>> takes significant time, and there aren't project-specific mirrors here. You
>>>>>>>>> might be able to point at a closer public mirror, depending on where you
>>>>>>>>> are.
>>>>>>>>>
>>>>>>>>> On Thu, Aug 15, 2019 at 5:43 AM Tianhua huang <
>>>>>>>>> huangtianhua223@gmail.com> wrote:
>>>>>>>>>
>>>>>>>>>> Hi all,
>>>>>>>>>>
>>>>>>>>>> I want to discuss spark ARM CI again, we took some tests on arm
>>>>>>>>>> instance based on master and the job includes
>>>>>>>>>> https://github.com/theopenlab/spark/pull/13 and k8s integration
>>>>>>>>>> https://github.com/theopenlab/spark/pull/17/ , there are several
>>>>>>>>>> things I want to talk about:
>>>>>>>>>>
>>>>>>>>>> First, about the failed tests:
>>>>>>>>>> 1.we have fixed some problems like
>>>>>>>>>> https://github.com/apache/spark/pull/25186 and
>>>>>>>>>> https://github.com/apache/spark/pull/25279, thanks sean owen and
>>>>>>>>>> others to help us.
>>>>>>>>>> 2.we tried k8s integration test on arm, and met an error: apk
>>>>>>>>>> fetch hangs, the tests passed after adding '--network host' option for
>>>>>>>>>> command `docker build`, see:
>>>>>>>>>>
>>>>>>>>>> https://github.com/theopenlab/spark/pull/17/files#diff-5b731b14068240d63a93c393f6f9b1e8R176
>>>>>>>>>> , the solution refers to
>>>>>>>>>> https://github.com/gliderlabs/docker-alpine/issues/307 and I
>>>>>>>>>> don't know whether it happened once in community CI, or maybe we should
>>>>>>>>>> submit a pr to pass '--network host' when `docker build`?
>>>>>>>>>> 3.we found there are two tests failed after the commit
>>>>>>>>>> https://github.com/apache/spark/pull/23767 :
>>>>>>>>>> ReplayListenerSuite:
>>>>>>>>>> - ...
>>>>>>>>>> - End-to-end replay *** FAILED ***
>>>>>>>>>> "[driver]" did not equal "[1]"
>>>>>>>>>> (JsonProtocolSuite.scala:622)
>>>>>>>>>> - End-to-end replay with compression *** FAILED ***
>>>>>>>>>> "[driver]" did not equal "[1]"
>>>>>>>>>> (JsonProtocolSuite.scala:622)
>>>>>>>>>>
>>>>>>>>>> we tried to revert the commit and then the tests passed,
>>>>>>>>>> the patch is too big and so sorry we can't find the reason till now, if you
>>>>>>>>>> are interesting please try it, and it will be very appreciate if
>>>>>>>>>> someone can help us to figure it out.
>>>>>>>>>>
>>>>>>>>>> Second, about the test time, we increased the flavor of arm
>>>>>>>>>> instance to 16U16G, but seems there was no significant improvement, the k8s
>>>>>>>>>> integration test took about one and a half hours, and the QA test(like
>>>>>>>>>> spark-master-test-maven-hadoop-2.7 community jenkins job) took about
>>>>>>>>>> seventeen hours(it is too long :(), we suspect that the reason is the
>>>>>>>>>> performance and network,
>>>>>>>>>> we split the jobs based on projects such as sql, core and so on,
>>>>>>>>>> the time can be decrease to about seven hours, see
>>>>>>>>>> https://github.com/theopenlab/spark/pull/19 We found the Spark
>>>>>>>>>> QA tests like
>>>>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/ ,
>>>>>>>>>> it looks all tests seem never download the jar packages from maven centry
>>>>>>>>>> repo(such as
>>>>>>>>>> https://repo.maven.apache.org/maven2/org/opencypher/okapi-api/0.4.2/okapi-api-0.4.2.jar).
>>>>>>>>>> So we want to know how the jenkins jobs can do that, is there a internal
>>>>>>>>>> maven repo launched? maybe we can do the same thing to avoid the network
>>>>>>>>>> connection cost during downloading the dependent jar packages.
>>>>>>>>>>
>>>>>>>>>> Third, the most important thing, it's about ARM CI of spark, we
>>>>>>>>>> believe that it is necessary, right? And you can see we really made a lot
>>>>>>>>>> of efforts, now the basic arm build/test jobs is ok, so we suggest to add
>>>>>>>>>> arm jobs to community, we can set them to novoting firstly, and
>>>>>>>>>> improve/rich the jobs step by step. Generally, there are two ways in our
>>>>>>>>>> mind to integrate the ARM CI for spark:
>>>>>>>>>> 1) We introduce openlab ARM CI into spark as a custom CI
>>>>>>>>>> system. We provide human resources and test ARM VMs, also we will focus on
>>>>>>>>>> the ARM related issues about Spark. We will push the PR into community.
>>>>>>>>>> 2) We donate ARM VM resources into existing amplab Jenkins.
>>>>>>>>>> We still provide human resources, focus on the ARM related issues about
>>>>>>>>>> Spark and push the PR into community.
>>>>>>>>>> Both options, we will provide human resources to maintain, of
>>>>>>>>>> course it will be great if we can work together. So please tell us which
>>>>>>>>>> option you would like? And let's move forward. Waiting for your reply,
>>>>>>>>>> thank you very much.
>>>>>>>>>>
>>>>>>>>>
>
> --
> Shane Knapp
> UC Berkeley EECS Research / RISELab Staff Technical Lead
> https://rise.cs.berkeley.edu
>
Re: Ask for ARM CI for spark
Posted by Shane Knapp <sk...@berkeley.edu>.
@Tianhua huang <hu...@gmail.com> sure, i think we can get
something sorted for the short-term.
all we need is ssh access (i can provide an ssh key), and i can then have
our jenkins master launch a remote worker on that instance.
instance setup, etc, will be up to you. my support for the time being will
be to create the job and 'best effort' for everything else.
this should get us up and running asap.
is there an open JIRA for jenkins/arm test support? we can move the
technical details about this idea there.
On Sun, Sep 15, 2019 at 9:03 PM Tianhua huang <hu...@gmail.com>
wrote:
> @Sean Owen <sr...@gmail.com> , so sorry to reply late, we had a
> Mid-Autumn holiday:)
>
> If you hope to integrate ARM CI to amplab jenkins, we can offer the arm
> instance, and then the ARM job will run together with other x86 jobs, so
> maybe there is a guideline to do this? @shane knapp <sk...@berkeley.edu>
> would you help us?
>
> On Thu, Sep 12, 2019 at 9:36 PM Sean Owen <sr...@gmail.com> wrote:
>
>> I don't know what's involved in actually accepting or operating those
>> machines, so can't comment there, but in the meantime it's good that you
>> are running these tests and can help report changes needed to keep it
>> working with ARM. I would continue with that for now.
>>
>> On Wed, Sep 11, 2019 at 10:06 PM Tianhua huang <hu...@gmail.com>
>> wrote:
>>
>>> Hi all,
>>>
>>> For the whole work process of spark ARM CI, we want to make 2 things
>>> clear.
>>>
>>> The first thing is:
>>> About spark ARM CI, now we have two periodic jobs, one job[1] based on
>>> commit[2](which already fixed the replay tests failed issue[3], we made a
>>> new test branch based on date 09-09-2019), the other job[4] based on spark
>>> master.
>>>
>>> The first job we test on the specified branch to prove that our ARM CI
>>> is good and stable.
>>> The second job checks spark master every day, then we can find whether
>>> the latest commits affect the ARM CI. According to the build history and
>>> result, it shows that some problems are easier to find on ARM like
>>> SPARK-28770 <https://issues.apache.org/jira/browse/SPARK-28770>, and it
>>> also shows that we would make efforts to trace and figure them out, till
>>> now we have found and fixed several problems[5][6][7], thanks everyone of
>>> the community :). And we believe that ARM CI is very necessary, right?
>>>
>>> The second thing is:
>>> We plan to run the jobs for a period of time, and you can see the result
>>> and logs from 'build history' of the jobs console, if everything goes well
>>> for one or two weeks could community accept the ARM CI? or how long the
>>> periodic jobs to run then our community could have enough confidence to
>>> accept the ARM CI? As you suggested before, it's good to integrate ARM CI
>>> to amplab jenkins, we agree that and we can donate the ARM instances and
>>> then maintain the ARM-related test jobs together with community, any
>>> thoughts?
>>>
>>> Thank you all!
>>>
>>> [1]
>>> http://status.openlabtesting.org/job/spark-unchanged-branch-unit-test-hadoop-2.7-arm64
>>> [2]
>>> https://github.com/apache/spark/commit/0ed9fae45769d4b06b8cf8128f462f09ff3d9a72
>>> [3] https://issues.apache.org/jira/browse/SPARK-28770
>>> [4]
>>> http://status.openlabtesting.org/builds?job_name=spark-master-unit-test-hadoop-2.7-arm64
>>> [5] https://github.com/apache/spark/pull/25186
>>> [6] https://github.com/apache/spark/pull/25279
>>> [7] https://github.com/apache/spark/pull/25673
>>>
>>>
>>>
>>> On Fri, Aug 16, 2019 at 11:24 PM Sean Owen <sr...@gmail.com> wrote:
>>>
>>>> Yes, I think it's just local caching. After you run the build you
>>>> should find lots of stuff cached at ~/.m2/repository and it won't download
>>>> every time.
>>>>
>>>> On Fri, Aug 16, 2019 at 3:01 AM bo zhaobo <bz...@gmail.com>
>>>> wrote:
>>>>
>>>>> Hi Sean,
>>>>> Thanks for reply. And very apologize for making you confused.
>>>>> I know the dependencies will be downloaded from SBT or Maven. But the
>>>>> Spark QA job also exec "mvn clean package", why the log didn't print
>>>>> "downloading some jar from Maven central [1] and build very fast. Is the
>>>>> reason that Spark Jenkins build the Spark jars in the physical machiines
>>>>> and won't destrory the test env after job is finished? Then the other job
>>>>> build Spark will get the dependencies jar from the local cached, as the
>>>>> previous jobs exec "mvn package", those dependencies had been downloaded
>>>>> already on local worker machine. Am I right? Is that the reason the job
>>>>> log[1] didn't print any downloading information from Maven Central?
>>>>>
>>>>> Thank you very much.
>>>>>
>>>>> [1]
>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/job/spark-master-test-maven-hadoop-2.6-ubuntu-testing/lastBuild/consoleFull
>>>>>
>>>>>
>>>>> Best regards
>>>>>
>>>>> ZhaoBo
>>>>>
>>>>> [image: Mailtrack]
>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>> notified by
>>>>> Mailtrack
>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/08/16
>>>>> 下午03:58:53
>>>>>
>>>>> Sean Owen <sr...@gmail.com> 于2019年8月16日周五 上午10:38写道：
>>>>>
>>>>>> I'm not sure what you mean. The dependencies are downloaded by SBT
>>>>>> and Maven like in any other project, and nothing about it is specific to
>>>>>> Spark.
>>>>>> The worker machines cache artifacts that are downloaded from these,
>>>>>> but this is a function of Maven and SBT, not Spark. You may find that the
>>>>>> initial download takes a long time.
>>>>>>
>>>>>> On Thu, Aug 15, 2019 at 9:02 PM bo zhaobo <
>>>>>> bzhaojyathousandy@gmail.com> wrote:
>>>>>>
>>>>>>> Hi Sean,
>>>>>>>
>>>>>>> Thanks very much for pointing out the roadmap. ;-). Then I think we
>>>>>>> will continue to focus on our test environment.
>>>>>>>
>>>>>>> For the networking problems, I mean that we can access Maven
>>>>>>> Central, and jobs cloud download the required jar package with a high
>>>>>>> network speed. What we want to know is that, why the Spark QA test jobs[1]
>>>>>>> log shows the job script/maven build seem don't download the jar packages?
>>>>>>> Could you tell us the reason about that? Thank you. The reason we raise
>>>>>>> the "networking problems" is that we found a phenomenon during we test, if
>>>>>>> we execute "mvn clean package" in a new test environment(As in our test
>>>>>>> environment, we will destory the test VMs after the job is finish), maven
>>>>>>> will download the dependency jar packages from Maven Central, but in this
>>>>>>> job "spark-master-test-maven-hadoop" [2], from the log, we didn't found it
>>>>>>> download any jar packages, what the reason about that?
>>>>>>> Also we build the Spark jar with downloading dependencies from Maven
>>>>>>> Central, it will cost mostly 1 hour. And we found [2] just cost 10min. But
>>>>>>> if we run "mvn package" in a VM which already exec "mvn package" before, it
>>>>>>> just cost 14min, looks very closer with [2]. So we suspect that downloading
>>>>>>> the Jar packages cost so much time. For the goad of ARM CI, we expect the
>>>>>>> performance of NEW ARM CI could be closer with existing X86 CI, then users
>>>>>>> could accept it eaiser.
>>>>>>>
>>>>>>> [1] https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/
>>>>>>> [2]
>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/job/spark-master-test-maven-hadoop-2.6-ubuntu-testing/lastBuild/consoleFull
>>>>>>>
>>>>>>> Best regards
>>>>>>>
>>>>>>> ZhaoBo
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> [image: Mailtrack]
>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>>> notified by
>>>>>>> Mailtrack
>>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/08/16
>>>>>>> 上午09:48:43
>>>>>>>
>>>>>>> Sean Owen <sr...@gmail.com> 于2019年8月15日周四 下午9:58写道：
>>>>>>>
>>>>>>>> I think the right goal is to fix the remaining issues first. If we
>>>>>>>> set up CI/CD it will only tell us there are still some test failures. If
>>>>>>>> it's stable, and not hard to add to the existing CI/CD, yes it could be
>>>>>>>> done automatically later. You can continue to test on ARM independently for
>>>>>>>> now.
>>>>>>>>
>>>>>>>> It sounds indeed like there are some networking problems in the
>>>>>>>> test system if you're not able to download from Maven Central. That rarely
>>>>>>>> takes significant time, and there aren't project-specific mirrors here. You
>>>>>>>> might be able to point at a closer public mirror, depending on where you
>>>>>>>> are.
>>>>>>>>
>>>>>>>> On Thu, Aug 15, 2019 at 5:43 AM Tianhua huang <
>>>>>>>> huangtianhua223@gmail.com> wrote:
>>>>>>>>
>>>>>>>>> Hi all,
>>>>>>>>>
>>>>>>>>> I want to discuss spark ARM CI again, we took some tests on arm
>>>>>>>>> instance based on master and the job includes
>>>>>>>>> https://github.com/theopenlab/spark/pull/13 and k8s integration
>>>>>>>>> https://github.com/theopenlab/spark/pull/17/ , there are several
>>>>>>>>> things I want to talk about:
>>>>>>>>>
>>>>>>>>> First, about the failed tests:
>>>>>>>>> 1.we have fixed some problems like
>>>>>>>>> https://github.com/apache/spark/pull/25186 and
>>>>>>>>> https://github.com/apache/spark/pull/25279, thanks sean owen and
>>>>>>>>> others to help us.
>>>>>>>>> 2.we tried k8s integration test on arm, and met an error: apk
>>>>>>>>> fetch hangs, the tests passed after adding '--network host' option for
>>>>>>>>> command `docker build`, see:
>>>>>>>>>
>>>>>>>>> https://github.com/theopenlab/spark/pull/17/files#diff-5b731b14068240d63a93c393f6f9b1e8R176
>>>>>>>>> , the solution refers to
>>>>>>>>> https://github.com/gliderlabs/docker-alpine/issues/307 and I
>>>>>>>>> don't know whether it happened once in community CI, or maybe we should
>>>>>>>>> submit a pr to pass '--network host' when `docker build`?
>>>>>>>>> 3.we found there are two tests failed after the commit
>>>>>>>>> https://github.com/apache/spark/pull/23767 :
>>>>>>>>> ReplayListenerSuite:
>>>>>>>>> - ...
>>>>>>>>> - End-to-end replay *** FAILED ***
>>>>>>>>> "[driver]" did not equal "[1]"
>>>>>>>>> (JsonProtocolSuite.scala:622)
>>>>>>>>> - End-to-end replay with compression *** FAILED ***
>>>>>>>>> "[driver]" did not equal "[1]"
>>>>>>>>> (JsonProtocolSuite.scala:622)
>>>>>>>>>
>>>>>>>>> we tried to revert the commit and then the tests passed,
>>>>>>>>> the patch is too big and so sorry we can't find the reason till now, if you
>>>>>>>>> are interesting please try it, and it will be very appreciate if
>>>>>>>>> someone can help us to figure it out.
>>>>>>>>>
>>>>>>>>> Second, about the test time, we increased the flavor of arm
>>>>>>>>> instance to 16U16G, but seems there was no significant improvement, the k8s
>>>>>>>>> integration test took about one and a half hours, and the QA test(like
>>>>>>>>> spark-master-test-maven-hadoop-2.7 community jenkins job) took about
>>>>>>>>> seventeen hours(it is too long :(), we suspect that the reason is the
>>>>>>>>> performance and network,
>>>>>>>>> we split the jobs based on projects such as sql, core and so on,
>>>>>>>>> the time can be decrease to about seven hours, see
>>>>>>>>> https://github.com/theopenlab/spark/pull/19 We found the Spark QA
>>>>>>>>> tests like
>>>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/ ,
>>>>>>>>> it looks all tests seem never download the jar packages from maven centry
>>>>>>>>> repo(such as
>>>>>>>>> https://repo.maven.apache.org/maven2/org/opencypher/okapi-api/0.4.2/okapi-api-0.4.2.jar).
>>>>>>>>> So we want to know how the jenkins jobs can do that, is there a internal
>>>>>>>>> maven repo launched? maybe we can do the same thing to avoid the network
>>>>>>>>> connection cost during downloading the dependent jar packages.
>>>>>>>>>
>>>>>>>>> Third, the most important thing, it's about ARM CI of spark, we
>>>>>>>>> believe that it is necessary, right? And you can see we really made a lot
>>>>>>>>> of efforts, now the basic arm build/test jobs is ok, so we suggest to add
>>>>>>>>> arm jobs to community, we can set them to novoting firstly, and
>>>>>>>>> improve/rich the jobs step by step. Generally, there are two ways in our
>>>>>>>>> mind to integrate the ARM CI for spark:
>>>>>>>>> 1) We introduce openlab ARM CI into spark as a custom CI
>>>>>>>>> system. We provide human resources and test ARM VMs, also we will focus on
>>>>>>>>> the ARM related issues about Spark. We will push the PR into community.
>>>>>>>>> 2) We donate ARM VM resources into existing amplab Jenkins.
>>>>>>>>> We still provide human resources, focus on the ARM related issues about
>>>>>>>>> Spark and push the PR into community.
>>>>>>>>> Both options, we will provide human resources to maintain, of
>>>>>>>>> course it will be great if we can work together. So please tell us which
>>>>>>>>> option you would like? And let's move forward. Waiting for your reply,
>>>>>>>>> thank you very much.
>>>>>>>>>
>>>>>>>>
--
Shane Knapp
UC Berkeley EECS Research / RISELab Staff Technical Lead
https://rise.cs.berkeley.edu
Re: Ask for ARM CI for spark
Posted by Tianhua huang <hu...@gmail.com>.
@Sean Owen <sr...@gmail.com> , so sorry to reply late, we had a Mid-Autumn
holiday:)
If you hope to integrate ARM CI to amplab jenkins, we can offer the arm
instance, and then the ARM job will run together with other x86 jobs, so
maybe there is a guideline to do this? @shane knapp <sk...@berkeley.edu>
would you help us?
On Thu, Sep 12, 2019 at 9:36 PM Sean Owen <sr...@gmail.com> wrote:
> I don't know what's involved in actually accepting or operating those
> machines, so can't comment there, but in the meantime it's good that you
> are running these tests and can help report changes needed to keep it
> working with ARM. I would continue with that for now.
>
> On Wed, Sep 11, 2019 at 10:06 PM Tianhua huang <hu...@gmail.com>
> wrote:
>
>> Hi all,
>>
>> For the whole work process of spark ARM CI, we want to make 2 things
>> clear.
>>
>> The first thing is:
>> About spark ARM CI, now we have two periodic jobs, one job[1] based on
>> commit[2](which already fixed the replay tests failed issue[3], we made a
>> new test branch based on date 09-09-2019), the other job[4] based on spark
>> master.
>>
>> The first job we test on the specified branch to prove that our ARM CI is
>> good and stable.
>> The second job checks spark master every day, then we can find whether
>> the latest commits affect the ARM CI. According to the build history and
>> result, it shows that some problems are easier to find on ARM like
>> SPARK-28770 <https://issues.apache.org/jira/browse/SPARK-28770>, and it
>> also shows that we would make efforts to trace and figure them out, till
>> now we have found and fixed several problems[5][6][7], thanks everyone of
>> the community :). And we believe that ARM CI is very necessary, right?
>>
>> The second thing is:
>> We plan to run the jobs for a period of time, and you can see the result
>> and logs from 'build history' of the jobs console, if everything goes well
>> for one or two weeks could community accept the ARM CI? or how long the
>> periodic jobs to run then our community could have enough confidence to
>> accept the ARM CI? As you suggested before, it's good to integrate ARM CI
>> to amplab jenkins, we agree that and we can donate the ARM instances and
>> then maintain the ARM-related test jobs together with community, any
>> thoughts?
>>
>> Thank you all!
>>
>> [1]
>> http://status.openlabtesting.org/job/spark-unchanged-branch-unit-test-hadoop-2.7-arm64
>> [2]
>> https://github.com/apache/spark/commit/0ed9fae45769d4b06b8cf8128f462f09ff3d9a72
>> [3] https://issues.apache.org/jira/browse/SPARK-28770
>> [4]
>> http://status.openlabtesting.org/builds?job_name=spark-master-unit-test-hadoop-2.7-arm64
>> [5] https://github.com/apache/spark/pull/25186
>> [6] https://github.com/apache/spark/pull/25279
>> [7] https://github.com/apache/spark/pull/25673
>>
>>
>>
>> On Fri, Aug 16, 2019 at 11:24 PM Sean Owen <sr...@gmail.com> wrote:
>>
>>> Yes, I think it's just local caching. After you run the build you should
>>> find lots of stuff cached at ~/.m2/repository and it won't download every
>>> time.
>>>
>>> On Fri, Aug 16, 2019 at 3:01 AM bo zhaobo <bz...@gmail.com>
>>> wrote:
>>>
>>>> Hi Sean,
>>>> Thanks for reply. And very apologize for making you confused.
>>>> I know the dependencies will be downloaded from SBT or Maven. But the
>>>> Spark QA job also exec "mvn clean package", why the log didn't print
>>>> "downloading some jar from Maven central [1] and build very fast. Is the
>>>> reason that Spark Jenkins build the Spark jars in the physical machiines
>>>> and won't destrory the test env after job is finished? Then the other job
>>>> build Spark will get the dependencies jar from the local cached, as the
>>>> previous jobs exec "mvn package", those dependencies had been downloaded
>>>> already on local worker machine. Am I right? Is that the reason the job
>>>> log[1] didn't print any downloading information from Maven Central?
>>>>
>>>> Thank you very much.
>>>>
>>>> [1]
>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/job/spark-master-test-maven-hadoop-2.6-ubuntu-testing/lastBuild/consoleFull
>>>>
>>>>
>>>> Best regards
>>>>
>>>> ZhaoBo
>>>>
>>>> [image: Mailtrack]
>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>> notified by
>>>> Mailtrack
>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/08/16
>>>> 下午03:58:53
>>>>
>>>> Sean Owen <sr...@gmail.com> 于2019年8月16日周五 上午10:38写道：
>>>>
>>>>> I'm not sure what you mean. The dependencies are downloaded by SBT and
>>>>> Maven like in any other project, and nothing about it is specific to Spark.
>>>>> The worker machines cache artifacts that are downloaded from these,
>>>>> but this is a function of Maven and SBT, not Spark. You may find that the
>>>>> initial download takes a long time.
>>>>>
>>>>> On Thu, Aug 15, 2019 at 9:02 PM bo zhaobo <bz...@gmail.com>
>>>>> wrote:
>>>>>
>>>>>> Hi Sean,
>>>>>>
>>>>>> Thanks very much for pointing out the roadmap. ;-). Then I think we
>>>>>> will continue to focus on our test environment.
>>>>>>
>>>>>> For the networking problems, I mean that we can access Maven Central,
>>>>>> and jobs cloud download the required jar package with a high network speed.
>>>>>> What we want to know is that, why the Spark QA test jobs[1] log shows the
>>>>>> job script/maven build seem don't download the jar packages? Could you tell
>>>>>> us the reason about that? Thank you. The reason we raise the "networking
>>>>>> problems" is that we found a phenomenon during we test, if we execute "mvn
>>>>>> clean package" in a new test environment(As in our test environment, we
>>>>>> will destory the test VMs after the job is finish), maven will download the
>>>>>> dependency jar packages from Maven Central, but in this job
>>>>>> "spark-master-test-maven-hadoop" [2], from the log, we didn't found it
>>>>>> download any jar packages, what the reason about that?
>>>>>> Also we build the Spark jar with downloading dependencies from Maven
>>>>>> Central, it will cost mostly 1 hour. And we found [2] just cost 10min. But
>>>>>> if we run "mvn package" in a VM which already exec "mvn package" before, it
>>>>>> just cost 14min, looks very closer with [2]. So we suspect that downloading
>>>>>> the Jar packages cost so much time. For the goad of ARM CI, we expect the
>>>>>> performance of NEW ARM CI could be closer with existing X86 CI, then users
>>>>>> could accept it eaiser.
>>>>>>
>>>>>> [1] https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/
>>>>>> [2]
>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/job/spark-master-test-maven-hadoop-2.6-ubuntu-testing/lastBuild/consoleFull
>>>>>>
>>>>>> Best regards
>>>>>>
>>>>>> ZhaoBo
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> [image: Mailtrack]
>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>>> notified by
>>>>>> Mailtrack
>>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/08/16
>>>>>> 上午09:48:43
>>>>>>
>>>>>> Sean Owen <sr...@gmail.com> 于2019年8月15日周四 下午9:58写道：
>>>>>>
>>>>>>> I think the right goal is to fix the remaining issues first. If we
>>>>>>> set up CI/CD it will only tell us there are still some test failures. If
>>>>>>> it's stable, and not hard to add to the existing CI/CD, yes it could be
>>>>>>> done automatically later. You can continue to test on ARM independently for
>>>>>>> now.
>>>>>>>
>>>>>>> It sounds indeed like there are some networking problems in the test
>>>>>>> system if you're not able to download from Maven Central. That rarely takes
>>>>>>> significant time, and there aren't project-specific mirrors here. You might
>>>>>>> be able to point at a closer public mirror, depending on where you are.
>>>>>>>
>>>>>>> On Thu, Aug 15, 2019 at 5:43 AM Tianhua huang <
>>>>>>> huangtianhua223@gmail.com> wrote:
>>>>>>>
>>>>>>>> Hi all,
>>>>>>>>
>>>>>>>> I want to discuss spark ARM CI again, we took some tests on arm
>>>>>>>> instance based on master and the job includes
>>>>>>>> https://github.com/theopenlab/spark/pull/13 and k8s integration
>>>>>>>> https://github.com/theopenlab/spark/pull/17/ , there are several
>>>>>>>> things I want to talk about:
>>>>>>>>
>>>>>>>> First, about the failed tests:
>>>>>>>> 1.we have fixed some problems like
>>>>>>>> https://github.com/apache/spark/pull/25186 and
>>>>>>>> https://github.com/apache/spark/pull/25279, thanks sean owen and
>>>>>>>> others to help us.
>>>>>>>> 2.we tried k8s integration test on arm, and met an error: apk
>>>>>>>> fetch hangs, the tests passed after adding '--network host' option for
>>>>>>>> command `docker build`, see:
>>>>>>>>
>>>>>>>> https://github.com/theopenlab/spark/pull/17/files#diff-5b731b14068240d63a93c393f6f9b1e8R176
>>>>>>>> , the solution refers to
>>>>>>>> https://github.com/gliderlabs/docker-alpine/issues/307 and I
>>>>>>>> don't know whether it happened once in community CI, or maybe we should
>>>>>>>> submit a pr to pass '--network host' when `docker build`?
>>>>>>>> 3.we found there are two tests failed after the commit
>>>>>>>> https://github.com/apache/spark/pull/23767 :
>>>>>>>> ReplayListenerSuite:
>>>>>>>> - ...
>>>>>>>> - End-to-end replay *** FAILED ***
>>>>>>>> "[driver]" did not equal "[1]"
>>>>>>>> (JsonProtocolSuite.scala:622)
>>>>>>>> - End-to-end replay with compression *** FAILED ***
>>>>>>>> "[driver]" did not equal "[1]"
>>>>>>>> (JsonProtocolSuite.scala:622)
>>>>>>>>
>>>>>>>> we tried to revert the commit and then the tests passed,
>>>>>>>> the patch is too big and so sorry we can't find the reason till now, if you
>>>>>>>> are interesting please try it, and it will be very appreciate if
>>>>>>>> someone can help us to figure it out.
>>>>>>>>
>>>>>>>> Second, about the test time, we increased the flavor of arm
>>>>>>>> instance to 16U16G, but seems there was no significant improvement, the k8s
>>>>>>>> integration test took about one and a half hours, and the QA test(like
>>>>>>>> spark-master-test-maven-hadoop-2.7 community jenkins job) took about
>>>>>>>> seventeen hours(it is too long :(), we suspect that the reason is the
>>>>>>>> performance and network,
>>>>>>>> we split the jobs based on projects such as sql, core and so on,
>>>>>>>> the time can be decrease to about seven hours, see
>>>>>>>> https://github.com/theopenlab/spark/pull/19 We found the Spark QA
>>>>>>>> tests like
>>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/ ,
>>>>>>>> it looks all tests seem never download the jar packages from maven centry
>>>>>>>> repo(such as
>>>>>>>> https://repo.maven.apache.org/maven2/org/opencypher/okapi-api/0.4.2/okapi-api-0.4.2.jar).
>>>>>>>> So we want to know how the jenkins jobs can do that, is there a internal
>>>>>>>> maven repo launched? maybe we can do the same thing to avoid the network
>>>>>>>> connection cost during downloading the dependent jar packages.
>>>>>>>>
>>>>>>>> Third, the most important thing, it's about ARM CI of spark, we
>>>>>>>> believe that it is necessary, right? And you can see we really made a lot
>>>>>>>> of efforts, now the basic arm build/test jobs is ok, so we suggest to add
>>>>>>>> arm jobs to community, we can set them to novoting firstly, and
>>>>>>>> improve/rich the jobs step by step. Generally, there are two ways in our
>>>>>>>> mind to integrate the ARM CI for spark:
>>>>>>>> 1) We introduce openlab ARM CI into spark as a custom CI
>>>>>>>> system. We provide human resources and test ARM VMs, also we will focus on
>>>>>>>> the ARM related issues about Spark. We will push the PR into community.
>>>>>>>> 2) We donate ARM VM resources into existing amplab Jenkins. We
>>>>>>>> still provide human resources, focus on the ARM related issues about Spark
>>>>>>>> and push the PR into community.
>>>>>>>> Both options, we will provide human resources to maintain, of
>>>>>>>> course it will be great if we can work together. So please tell us which
>>>>>>>> option you would like? And let's move forward. Waiting for your reply,
>>>>>>>> thank you very much.
>>>>>>>>
>>>>>>>
Re: Ask for ARM CI for spark
Posted by Sean Owen <sr...@gmail.com>.
I don't know what's involved in actually accepting or operating those
machines, so can't comment there, but in the meantime it's good that you
are running these tests and can help report changes needed to keep it
working with ARM. I would continue with that for now.
On Wed, Sep 11, 2019 at 10:06 PM Tianhua huang <hu...@gmail.com>
wrote:
> Hi all,
>
> For the whole work process of spark ARM CI, we want to make 2 things clear.
>
> The first thing is:
> About spark ARM CI, now we have two periodic jobs, one job[1] based on
> commit[2](which already fixed the replay tests failed issue[3], we made a
> new test branch based on date 09-09-2019), the other job[4] based on spark
> master.
>
> The first job we test on the specified branch to prove that our ARM CI is
> good and stable.
> The second job checks spark master every day, then we can find whether the
> latest commits affect the ARM CI. According to the build history and
> result, it shows that some problems are easier to find on ARM like
> SPARK-28770 <https://issues.apache.org/jira/browse/SPARK-28770>, and it
> also shows that we would make efforts to trace and figure them out, till
> now we have found and fixed several problems[5][6][7], thanks everyone of
> the community :). And we believe that ARM CI is very necessary, right?
>
> The second thing is:
> We plan to run the jobs for a period of time, and you can see the result
> and logs from 'build history' of the jobs console, if everything goes well
> for one or two weeks could community accept the ARM CI? or how long the
> periodic jobs to run then our community could have enough confidence to
> accept the ARM CI? As you suggested before, it's good to integrate ARM CI
> to amplab jenkins, we agree that and we can donate the ARM instances and
> then maintain the ARM-related test jobs together with community, any
> thoughts?
>
> Thank you all!
>
> [1]
> http://status.openlabtesting.org/job/spark-unchanged-branch-unit-test-hadoop-2.7-arm64
> [2]
> https://github.com/apache/spark/commit/0ed9fae45769d4b06b8cf8128f462f09ff3d9a72
> [3] https://issues.apache.org/jira/browse/SPARK-28770
> [4]
> http://status.openlabtesting.org/builds?job_name=spark-master-unit-test-hadoop-2.7-arm64
> [5] https://github.com/apache/spark/pull/25186
> [6] https://github.com/apache/spark/pull/25279
> [7] https://github.com/apache/spark/pull/25673
>
>
>
> On Fri, Aug 16, 2019 at 11:24 PM Sean Owen <sr...@gmail.com> wrote:
>
>> Yes, I think it's just local caching. After you run the build you should
>> find lots of stuff cached at ~/.m2/repository and it won't download every
>> time.
>>
>> On Fri, Aug 16, 2019 at 3:01 AM bo zhaobo <bz...@gmail.com>
>> wrote:
>>
>>> Hi Sean,
>>> Thanks for reply. And very apologize for making you confused.
>>> I know the dependencies will be downloaded from SBT or Maven. But the
>>> Spark QA job also exec "mvn clean package", why the log didn't print
>>> "downloading some jar from Maven central [1] and build very fast. Is the
>>> reason that Spark Jenkins build the Spark jars in the physical machiines
>>> and won't destrory the test env after job is finished? Then the other job
>>> build Spark will get the dependencies jar from the local cached, as the
>>> previous jobs exec "mvn package", those dependencies had been downloaded
>>> already on local worker machine. Am I right? Is that the reason the job
>>> log[1] didn't print any downloading information from Maven Central?
>>>
>>> Thank you very much.
>>>
>>> [1]
>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/job/spark-master-test-maven-hadoop-2.6-ubuntu-testing/lastBuild/consoleFull
>>>
>>>
>>> Best regards
>>>
>>> ZhaoBo
>>>
>>> [image: Mailtrack]
>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>> notified by
>>> Mailtrack
>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/08/16
>>> 下午03:58:53
>>>
>>> Sean Owen <sr...@gmail.com> 于2019年8月16日周五 上午10:38写道：
>>>
>>>> I'm not sure what you mean. The dependencies are downloaded by SBT and
>>>> Maven like in any other project, and nothing about it is specific to Spark.
>>>> The worker machines cache artifacts that are downloaded from these, but
>>>> this is a function of Maven and SBT, not Spark. You may find that the
>>>> initial download takes a long time.
>>>>
>>>> On Thu, Aug 15, 2019 at 9:02 PM bo zhaobo <bz...@gmail.com>
>>>> wrote:
>>>>
>>>>> Hi Sean,
>>>>>
>>>>> Thanks very much for pointing out the roadmap. ;-). Then I think we
>>>>> will continue to focus on our test environment.
>>>>>
>>>>> For the networking problems, I mean that we can access Maven Central,
>>>>> and jobs cloud download the required jar package with a high network speed.
>>>>> What we want to know is that, why the Spark QA test jobs[1] log shows the
>>>>> job script/maven build seem don't download the jar packages? Could you tell
>>>>> us the reason about that? Thank you. The reason we raise the "networking
>>>>> problems" is that we found a phenomenon during we test, if we execute "mvn
>>>>> clean package" in a new test environment(As in our test environment, we
>>>>> will destory the test VMs after the job is finish), maven will download the
>>>>> dependency jar packages from Maven Central, but in this job
>>>>> "spark-master-test-maven-hadoop" [2], from the log, we didn't found it
>>>>> download any jar packages, what the reason about that?
>>>>> Also we build the Spark jar with downloading dependencies from Maven
>>>>> Central, it will cost mostly 1 hour. And we found [2] just cost 10min. But
>>>>> if we run "mvn package" in a VM which already exec "mvn package" before, it
>>>>> just cost 14min, looks very closer with [2]. So we suspect that downloading
>>>>> the Jar packages cost so much time. For the goad of ARM CI, we expect the
>>>>> performance of NEW ARM CI could be closer with existing X86 CI, then users
>>>>> could accept it eaiser.
>>>>>
>>>>> [1] https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/
>>>>> [2]
>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/job/spark-master-test-maven-hadoop-2.6-ubuntu-testing/lastBuild/consoleFull
>>>>>
>>>>> Best regards
>>>>>
>>>>> ZhaoBo
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> [image: Mailtrack]
>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
>>>>> notified by
>>>>> Mailtrack
>>>>> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 19/08/16
>>>>> 上午09:48:43
>>>>>
>>>>> Sean Owen <sr...@gmail.com> 于2019年8月15日周四 下午9:58写道：
>>>>>
>>>>>> I think the right goal is to fix the remaining issues first. If we
>>>>>> set up CI/CD it will only tell us there are still some test failures. If
>>>>>> it's stable, and not hard to add to the existing CI/CD, yes it could be
>>>>>> done automatically later. You can continue to test on ARM independently for
>>>>>> now.
>>>>>>
>>>>>> It sounds indeed like there are some networking problems in the test
>>>>>> system if you're not able to download from Maven Central. That rarely takes
>>>>>> significant time, and there aren't project-specific mirrors here. You might
>>>>>> be able to point at a closer public mirror, depending on where you are.
>>>>>>
>>>>>> On Thu, Aug 15, 2019 at 5:43 AM Tianhua huang <
>>>>>> huangtianhua223@gmail.com> wrote:
>>>>>>
>>>>>>> Hi all,
>>>>>>>
>>>>>>> I want to discuss spark ARM CI again, we took some tests on arm
>>>>>>> instance based on master and the job includes
>>>>>>> https://github.com/theopenlab/spark/pull/13 and k8s integration
>>>>>>> https://github.com/theopenlab/spark/pull/17/ , there are several
>>>>>>> things I want to talk about:
>>>>>>>
>>>>>>> First, about the failed tests:
>>>>>>> 1.we have fixed some problems like
>>>>>>> https://github.com/apache/spark/pull/25186 and
>>>>>>> https://github.com/apache/spark/pull/25279, thanks sean owen and
>>>>>>> others to help us.
>>>>>>> 2.we tried k8s integration test on arm, and met an error: apk
>>>>>>> fetch hangs, the tests passed after adding '--network host' option for
>>>>>>> command `docker build`, see:
>>>>>>>
>>>>>>> https://github.com/theopenlab/spark/pull/17/files#diff-5b731b14068240d63a93c393f6f9b1e8R176
>>>>>>> , the solution refers to
>>>>>>> https://github.com/gliderlabs/docker-alpine/issues/307 and I don't
>>>>>>> know whether it happened once in community CI, or maybe we should submit a
>>>>>>> pr to pass '--network host' when `docker build`?
>>>>>>> 3.we found there are two tests failed after the commit
>>>>>>> https://github.com/apache/spark/pull/23767 :
>>>>>>> ReplayListenerSuite:
>>>>>>> - ...
>>>>>>> - End-to-end replay *** FAILED ***
>>>>>>> "[driver]" did not equal "[1]" (JsonProtocolSuite.scala:622)
>>>>>>> - End-to-end replay with compression *** FAILED ***
>>>>>>> "[driver]" did not equal "[1]" (JsonProtocolSuite.scala:622)
>>>>>>>
>>>>>>> we tried to revert the commit and then the tests passed, the
>>>>>>> patch is too big and so sorry we can't find the reason till now, if you are
>>>>>>> interesting please try it, and it will be very appreciate if
>>>>>>> someone can help us to figure it out.
>>>>>>>
>>>>>>> Second, about the test time, we increased the flavor of arm instance
>>>>>>> to 16U16G, but seems there was no significant improvement, the k8s
>>>>>>> integration test took about one and a half hours, and the QA test(like
>>>>>>> spark-master-test-maven-hadoop-2.7 community jenkins job) took about
>>>>>>> seventeen hours(it is too long :(), we suspect that the reason is the
>>>>>>> performance and network,
>>>>>>> we split the jobs based on projects such as sql, core and so on, the
>>>>>>> time can be decrease to about seven hours, see
>>>>>>> https://github.com/theopenlab/spark/pull/19 We found the Spark QA
>>>>>>> tests like
>>>>>>> https://amplab.cs.berkeley.edu/jenkins/view/Spark%20QA%20Test/ ,
>>>>>>> it looks all tests seem never download the jar packages from maven centry
>>>>>>> repo(such as
>>>>>>> https://repo.maven.apache.org/maven2/org/opencypher/okapi-api/0.4.2/okapi-api-0.4.2.jar).
>>>>>>> So we want to know how the jenkins jobs can do that, is there a internal
>>>>>>> maven repo launched? maybe we can do the same thing to avoid the network
>>>>>>> connection cost during downloading the dependent jar packages.
>>>>>>>
>>>>>>> Third, the most important thing, it's about ARM CI of spark, we
>>>>>>> believe that it is necessary, right? And you can see we really made a lot
>>>>>>> of efforts, now the basic arm build/test jobs is ok, so we suggest to add
>>>>>>> arm jobs to community, we can set them to novoting firstly, and
>>>>>>> improve/rich the jobs step by step. Generally, there are two ways in our
>>>>>>> mind to integrate the ARM CI for spark:
>>>>>>> 1) We introduce openlab ARM CI into spark as a custom CI
>>>>>>> system. We provide human resources and test ARM VMs, also we will focus on
>>>>>>> the ARM related issues about Spark. We will push the PR into community.
>>>>>>> 2) We donate ARM VM resources into existing amplab Jenkins. We
>>>>>>> still provide human resources, focus on the ARM related issues about Spark
>>>>>>> and push the PR into community.
>>>>>>> Both options, we will provide human resources to maintain, of course
>>>>>>> it will be great if we can work together. So please tell us which option
>>>>>>> you would like? And let's move forward. Waiting for your reply, thank you
>>>>>>> very much.
>>>>>>>
>>>>>>