You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@ignite.apache.org by Maxim Muzafarov <ma...@gmail.com> on 2018/02/26 10:40:23 UTC
Test testJobIdCollision to use multiple JVMs [IGNITE-4706]
Hi all,
I'm triyng to clarify for myseft issue [1] of rewriting this test case to
use multiple JVMs. I'm trying to reproduce it using steps described here
[2]:
As I correct understand issue description, I'm runing testJobIdCollision
and expecting to get exception:
"Received computation request with duplicate job ID"
, but I've got:
"Job has already been hold [ctx=GridJobContextImpl
[jobId=f7e74a1d161-08edbe47-9b65-4ed2-8d0c-a8a1a6700003, timeoutObj=null,
attrs={}]]"
So, does this test-case actual or can be removed? Or we should use
another IgniteCallable<Object>
like othis one: IgniteWalRecoveryTest.LoadRunnable [4]?
Also, IgniteClusterProcessProxy#forNodeId [3] doesn't implemented yet.
Brief search for some JIRA's of implementation this method doesn't return
anything.
What should we do with this?
[1] https://issues.apache.org/jira/browse/IGNITE-4706
[2] https://issues.apache.org/jira/browse/IGNITE-1384
[3]
https://github.com/apache/ignite/blob/master/modules/core/src/test/java/org/apache/ignite/testframework/junits/multijvm/IgniteClusterProcessProxy.java#L204
[4]
https://github.com/apache/ignite/blob/master/modules/core/src/test/java/org/apache/ignite/internal/processors/cache/persistence/db/wal/IgniteWalRecoveryTest.java#L1552
Re: Test testJobIdCollision to use multiple JVMs [IGNITE-4706]
Posted by Maxim Muzafarov <ma...@gmail.com>.
Sergey, Dmirty,
Thanks for you attention.
Add comment to JIRA "why this task became minor priority" as mentioned
Sergey.
I'll try to find for mysefl more valuable for community.
пт, 16 мар. 2018 г. в 19:12, Dmitry Pavlov <dp...@gmail.com>:
> Hi Sergey,
>
> Thank you for stepping in.
>
> There is fresh test failures scope reflected in JIRA
>
> https://issues.apache.org/jira/issues/?jql=project%20%3D%20IGNITE%20AND%20labels%20%3D%20MakeTeamcityGreenAgain%20AND%20assignee%20is%20EMPTY%20AND%20resolution%20is%20EMPTY%20order%20by%20createdDate%20DESC%20%20%20
>
> And any test failure from CI without investigation also may be picked up.
>
> Sincerely,
> Dmitriy Pavlov
>
>
>
> пт, 16 мар. 2018 г. в 18:47, Sergey Chugunov <se...@gmail.com>:
>
> > Dmitry, Maxim,
> >
> > Thanks for bringing this up.
> >
> > I reviewed all context about [1], it looks like the test is still valid
> > but is of low priority; I reflected it in jira ticket itself.
> >
> > Also rewriting the test to multi-JVM fashion isn't an easy task, to me it
> > is much better to spend this time working on more important stuff.
> > Dmitry, could you suggest a better ticket for Maxim to look into?
> >
> > [1] https://issues.apache.org/jira/browse/IGNITE-4706
> >
> > --
> > Thanks,
> > Sergey.
> >
> > On Fri, Mar 16, 2018 at 4:43 PM, Dmitry Pavlov <dp...@gmail.com>
> > wrote:
> >
> >> Hi Sergey,
> >>
> >> Is this issue still actual for you?
> >>
> >> Sincerely,
> >> Dmitriy Pavlov
> >>
> >> пн, 26 февр. 2018 г. в 13:40, Maxim Muzafarov <ma...@gmail.com>:
> >>
> >>> Hi all,
> >>>
> >>> I'm triyng to clarify for myseft issue [1] of rewriting this test case
> to
> >>> use multiple JVMs. I'm trying to reproduce it using steps described
> here
> >>> [2]:
> >>> As I correct understand issue description, I'm runing
> testJobIdCollision
> >>> and expecting to get exception:
> >>> "Received computation request with duplicate job ID"
> >>> , but I've got:
> >>> "Job has already been hold [ctx=GridJobContextImpl
> >>> [jobId=f7e74a1d161-08edbe47-9b65-4ed2-8d0c-a8a1a6700003,
> timeoutObj=null,
> >>> attrs={}]]"
> >>>
> >>> So, does this test-case actual or can be removed? Or we should use
> >>> another IgniteCallable<Object>
> >>> like othis one: IgniteWalRecoveryTest.LoadRunnable [4]?
> >>>
> >>> Also, IgniteClusterProcessProxy#forNodeId [3] doesn't implemented yet.
> >>> Brief search for some JIRA's of implementation this method doesn't
> return
> >>> anything.
> >>> What should we do with this?
> >>>
> >>>
> >>>
> >>> [1] https://issues.apache.org/jira/browse/IGNITE-4706
> >>> [2] https://issues.apache.org/jira/browse/IGNITE-1384
> >>> [3]
> >>>
> >>>
> https://github.com/apache/ignite/blob/master/modules/core/src/test/java/org/apache/ignite/testframework/junits/multijvm/IgniteClusterProcessProxy.java#L204
> >>> [4]
> >>>
> >>>
> https://github.com/apache/ignite/blob/master/modules/core/src/test/java/org/apache/ignite/internal/processors/cache/persistence/db/wal/IgniteWalRecoveryTest.java#L1552
> >>>
> >>
> >
>
Re: Test testJobIdCollision to use multiple JVMs [IGNITE-4706]
Posted by Dmitry Pavlov <dp...@gmail.com>.
Hi Sergey,
Thank you for stepping in.
There is fresh test failures scope reflected in JIRA
https://issues.apache.org/jira/issues/?jql=project%20%3D%20IGNITE%20AND%20labels%20%3D%20MakeTeamcityGreenAgain%20AND%20assignee%20is%20EMPTY%20AND%20resolution%20is%20EMPTY%20order%20by%20createdDate%20DESC%20%20%20
And any test failure from CI without investigation also may be picked up.
Sincerely,
Dmitriy Pavlov
пт, 16 мар. 2018 г. в 18:47, Sergey Chugunov <se...@gmail.com>:
> Dmitry, Maxim,
>
> Thanks for bringing this up.
>
> I reviewed all context about [1], it looks like the test is still valid
> but is of low priority; I reflected it in jira ticket itself.
>
> Also rewriting the test to multi-JVM fashion isn't an easy task, to me it
> is much better to spend this time working on more important stuff.
> Dmitry, could you suggest a better ticket for Maxim to look into?
>
> [1] https://issues.apache.org/jira/browse/IGNITE-4706
>
> --
> Thanks,
> Sergey.
>
> On Fri, Mar 16, 2018 at 4:43 PM, Dmitry Pavlov <dp...@gmail.com>
> wrote:
>
>> Hi Sergey,
>>
>> Is this issue still actual for you?
>>
>> Sincerely,
>> Dmitriy Pavlov
>>
>> пн, 26 февр. 2018 г. в 13:40, Maxim Muzafarov <ma...@gmail.com>:
>>
>>> Hi all,
>>>
>>> I'm triyng to clarify for myseft issue [1] of rewriting this test case to
>>> use multiple JVMs. I'm trying to reproduce it using steps described here
>>> [2]:
>>> As I correct understand issue description, I'm runing testJobIdCollision
>>> and expecting to get exception:
>>> "Received computation request with duplicate job ID"
>>> , but I've got:
>>> "Job has already been hold [ctx=GridJobContextImpl
>>> [jobId=f7e74a1d161-08edbe47-9b65-4ed2-8d0c-a8a1a6700003, timeoutObj=null,
>>> attrs={}]]"
>>>
>>> So, does this test-case actual or can be removed? Or we should use
>>> another IgniteCallable<Object>
>>> like othis one: IgniteWalRecoveryTest.LoadRunnable [4]?
>>>
>>> Also, IgniteClusterProcessProxy#forNodeId [3] doesn't implemented yet.
>>> Brief search for some JIRA's of implementation this method doesn't return
>>> anything.
>>> What should we do with this?
>>>
>>>
>>>
>>> [1] https://issues.apache.org/jira/browse/IGNITE-4706
>>> [2] https://issues.apache.org/jira/browse/IGNITE-1384
>>> [3]
>>>
>>> https://github.com/apache/ignite/blob/master/modules/core/src/test/java/org/apache/ignite/testframework/junits/multijvm/IgniteClusterProcessProxy.java#L204
>>> [4]
>>>
>>> https://github.com/apache/ignite/blob/master/modules/core/src/test/java/org/apache/ignite/internal/processors/cache/persistence/db/wal/IgniteWalRecoveryTest.java#L1552
>>>
>>
>
Re: Test testJobIdCollision to use multiple JVMs [IGNITE-4706]
Posted by Sergey Chugunov <se...@gmail.com>.
Dmitry, Maxim,
Thanks for bringing this up.
I reviewed all context about [1], it looks like the test is still valid but
is of low priority; I reflected it in jira ticket itself.
Also rewriting the test to multi-JVM fashion isn't an easy task, to me it
is much better to spend this time working on more important stuff.
Dmitry, could you suggest a better ticket for Maxim to look into?
[1] https://issues.apache.org/jira/browse/IGNITE-4706
--
Thanks,
Sergey.
On Fri, Mar 16, 2018 at 4:43 PM, Dmitry Pavlov <dp...@gmail.com>
wrote:
> Hi Sergey,
>
> Is this issue still actual for you?
>
> Sincerely,
> Dmitriy Pavlov
>
> пн, 26 февр. 2018 г. в 13:40, Maxim Muzafarov <ma...@gmail.com>:
>
>> Hi all,
>>
>> I'm triyng to clarify for myseft issue [1] of rewriting this test case to
>> use multiple JVMs. I'm trying to reproduce it using steps described here
>> [2]:
>> As I correct understand issue description, I'm runing testJobIdCollision
>> and expecting to get exception:
>> "Received computation request with duplicate job ID"
>> , but I've got:
>> "Job has already been hold [ctx=GridJobContextImpl
>> [jobId=f7e74a1d161-08edbe47-9b65-4ed2-8d0c-a8a1a6700003, timeoutObj=null,
>> attrs={}]]"
>>
>> So, does this test-case actual or can be removed? Or we should use
>> another IgniteCallable<Object>
>> like othis one: IgniteWalRecoveryTest.LoadRunnable [4]?
>>
>> Also, IgniteClusterProcessProxy#forNodeId [3] doesn't implemented yet.
>> Brief search for some JIRA's of implementation this method doesn't return
>> anything.
>> What should we do with this?
>>
>>
>>
>> [1] https://issues.apache.org/jira/browse/IGNITE-4706
>> [2] https://issues.apache.org/jira/browse/IGNITE-1384
>> [3]
>> https://github.com/apache/ignite/blob/master/modules/
>> core/src/test/java/org/apache/ignite/testframework/junits/multijvm/
>> IgniteClusterProcessProxy.java#L204
>> [4]
>> https://github.com/apache/ignite/blob/master/modules/
>> core/src/test/java/org/apache/ignite/internal/processors/
>> cache/persistence/db/wal/IgniteWalRecoveryTest.java#L1552
>>
>
Re: Test testJobIdCollision to use multiple JVMs [IGNITE-4706]
Posted by Dmitry Pavlov <dp...@gmail.com>.
Hi Sergey,
Is this issue still actual for you?
Sincerely,
Dmitriy Pavlov
пн, 26 февр. 2018 г. в 13:40, Maxim Muzafarov <ma...@gmail.com>:
> Hi all,
>
> I'm triyng to clarify for myseft issue [1] of rewriting this test case to
> use multiple JVMs. I'm trying to reproduce it using steps described here
> [2]:
> As I correct understand issue description, I'm runing testJobIdCollision
> and expecting to get exception:
> "Received computation request with duplicate job ID"
> , but I've got:
> "Job has already been hold [ctx=GridJobContextImpl
> [jobId=f7e74a1d161-08edbe47-9b65-4ed2-8d0c-a8a1a6700003, timeoutObj=null,
> attrs={}]]"
>
> So, does this test-case actual or can be removed? Or we should use
> another IgniteCallable<Object>
> like othis one: IgniteWalRecoveryTest.LoadRunnable [4]?
>
> Also, IgniteClusterProcessProxy#forNodeId [3] doesn't implemented yet.
> Brief search for some JIRA's of implementation this method doesn't return
> anything.
> What should we do with this?
>
>
>
> [1] https://issues.apache.org/jira/browse/IGNITE-4706
> [2] https://issues.apache.org/jira/browse/IGNITE-1384
> [3]
>
> https://github.com/apache/ignite/blob/master/modules/core/src/test/java/org/apache/ignite/testframework/junits/multijvm/IgniteClusterProcessProxy.java#L204
> [4]
>
> https://github.com/apache/ignite/blob/master/modules/core/src/test/java/org/apache/ignite/internal/processors/cache/persistence/db/wal/IgniteWalRecoveryTest.java#L1552
>