You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@openwhisk.apache.org by David P Grove <gr...@us.ibm.com> on 2019/05/28 16:03:04 UTC
apigateway TravisCI test failure on master for almost 3 weeks
Hi,
It's been 20 days since the daily test run for apigateway has
actually succeeded. Spot-checking a few of the runs, it seems like a
consistent failure of one test case:
apigw.healthtests.ApiGwRestEndToEndTests > Wsk api should create an API and
successfully invoke that API FAILED
org.scalatest.exceptions.TestFailedException: 502 was not equal to 200
Does anyone have any insight into why this test is now failing?
thanks,
--dave
Re: apigateway TravisCI test failure on master for almost 3 weeks
Posted by Lars Andersson <la...@gmail.com>.
Thanks Rodric, I'll have a look tomorrow.
/ Lars
On Tue, 28 May 2019, 20:25 Rodric Rabbah <rodric@gmail.com wrote:
> I created an issue for this test failure here:
> https://github.com/apache/incubator-openwhisk-apigateway/issues/342
>
> Which implicates this line:
> tenants.lua:130: attempt to compare nil with number
>
>
> https://github.com/apache/incubator-openwhisk-apigateway/blob/1f46de9d2f7a9420e48c533af7a4b1b6118aa130/scripts/lua/management/lib/tenants.lua#L129-L130
>
> Lars - maybe you can spot the issue from that? I suspect guarding against
> nil is sufficient.
>
> -r
>
>
> On Tue, May 28, 2019 at 1:44 PM Lars Andersson <
> lars.storfiskarn.andersson@gmail.com> wrote:
>
> > If that PR is the cause, feel free to revert it as I have no time
> available
> > right now to have a look.
> >
> > / Lars
> >
> > On Tue, 28 May 2019, 19:35 Matt Hamann <matthew.hamann@gmail.com wrote:
> >
> > > Looks like the failures began directly after the merging of PR #335.
> > Rodric
> > > and Lars were the contributors of that PR.
> > >
> > > I would elect to either revert the PR or ask one of the aforementioned
> to
> > > investigate the failure and submit a fix. The cause of the failure is
> not
> > > obvious from the build output.
> > >
> > > ------------------------------------------------
> > > -Matt
> > > matthew.hamann@gmail.com
> > >
> > >
> > > On Tue, May 28, 2019 at 12:03 PM David P Grove <gr...@us.ibm.com>
> > wrote:
> > >
> > > >
> > > > Hi,
> > > >
> > > > It's been 20 days since the daily test run for apigateway has
> > > > actually succeeded. Spot-checking a few of the runs, it seems like a
> > > > consistent failure of one test case:
> > > >
> > > > apigw.healthtests.ApiGwRestEndToEndTests > Wsk api should create an
> API
> > > and
> > > > successfully invoke that API FAILED
> > > > org.scalatest.exceptions.TestFailedException: 502 was not equal
> to
> > > 200
> > > >
> > > > Does anyone have any insight into why this test is now
> failing?
> > > >
> > > > thanks,
> > > >
> > > > --dave
> > > >
> > >
> >
>
Re: apigateway TravisCI test failure on master for almost 3 weeks
Posted by Lars Andersson <la...@gmail.com>.
Rodric's patch looks good to me! I prefer the fix he made on line 117.
/ Lars
Den tis 28 maj 2019 kl 23:10 skrev David P Grove <gr...@us.ibm.com>:
> Thanks Rodric,
>
> Since this a regression vs. the previous release, I plan to cancel the
> current IPMC vote on releasing rc1 of apigateway 0.10.0-incubating and we
> will start fresh with a new release candidate after a fix is merged.
>
> --dave
>
>
> [image: Inactive hide details for Rodric Rabbah ---05/28/2019 03:22:14
> PM---I posted an explanation to the issue and opened a suggested]Rodric
> Rabbah ---05/28/2019 03:22:14 PM---I posted an explanation to the issue and
> opened a suggested patch https://urldefense.proofpoint.com/
>
> From: Rodric Rabbah <ro...@gmail.com>
> To: dev@openwhisk.apache.org
> Date: 05/28/2019 03:22 PM
> Subject: [EXTERNAL] Re: apigateway TravisCI test failure on master for
> almost 3 weeks
> ------------------------------
>
>
>
> I posted an explanation to the issue and opened a suggested patch
> https://github.com/apache/incubator-openwhisk-apigateway/pull/343.
>
> -r
>
> On Tue, May 28, 2019 at 2:25 PM Rodric Rabbah <ro...@gmail.com> wrote:
>
> > I created an issue for this test failure here:
> > https://github.com/apache/incubator-openwhisk-apigateway/issues/342
> >
> > Which implicates this line:
> > tenants.lua:130: attempt to compare nil with number
> >
> >
> >
> https://github.com/apache/incubator-openwhisk-apigateway/blob/1f46de9d2f7a9420e48c533af7a4b1b6118aa130/scripts/lua/management/lib/tenants.lua#L129-L130
> >
> > Lars - maybe you can spot the issue from that? I suspect guarding against
> > nil is sufficient.
> >
> > -r
> >
> >
> > On Tue, May 28, 2019 at 1:44 PM Lars Andersson <
> > lars.storfiskarn.andersson@gmail.com> wrote:
> >
> >> If that PR is the cause, feel free to revert it as I have no time
> >> available
> >> right now to have a look.
> >>
> >> / Lars
> >>
> >> On Tue, 28 May 2019, 19:35 Matt Hamann <matthew.hamann@gmail.com wrote:
> >>
> >> > Looks like the failures began directly after the merging of PR #335.
> >> Rodric
> >> > and Lars were the contributors of that PR.
> >> >
> >> > I would elect to either revert the PR or ask one of the aforementioned
> >> to
> >> > investigate the failure and submit a fix. The cause of the failure is
> >> not
> >> > obvious from the build output.
> >> >
> >> > ------------------------------------------------
> >> > -Matt
> >> > matthew.hamann@gmail.com
> >> >
> >> >
> >> > On Tue, May 28, 2019 at 12:03 PM David P Grove <gr...@us.ibm.com>
> >> wrote:
> >> >
> >> > >
> >> > > Hi,
> >> > >
> >> > > It's been 20 days since the daily test run for apigateway
> has
> >> > > actually succeeded. Spot-checking a few of the runs, it seems like
> a
> >> > > consistent failure of one test case:
> >> > >
> >> > > apigw.healthtests.ApiGwRestEndToEndTests > Wsk api should create an
> >> API
> >> > and
> >> > > successfully invoke that API FAILED
> >> > > org.scalatest.exceptions.TestFailedException: 502 was not equal
> to
> >> > 200
> >> > >
> >> > > Does anyone have any insight into why this test is now
> >> failing?
> >> > >
> >> > > thanks,
> >> > >
> >> > > --dave
> >> > >
> >> >
> >>
> >
>
>
>
>
Re: apigateway TravisCI test failure on master for almost 3 weeks
Posted by David P Grove <gr...@us.ibm.com>.
Thanks Rodric,
Since this a regression vs. the previous release, I plan to cancel the
current IPMC vote on releasing rc1 of apigateway 0.10.0-incubating and we
will start fresh with a new release candidate after a fix is merged.
--dave
From: Rodric Rabbah <ro...@gmail.com>
To: dev@openwhisk.apache.org
Date: 05/28/2019 03:22 PM
Subject: [EXTERNAL] Re: apigateway TravisCI test failure on master for
almost 3 weeks
I posted an explanation to the issue and opened a suggested patch
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_apache_incubator-2Dopenwhisk-2Dapigateway_pull_343&d=DwIBaQ&c=jf_iaSHvJObTbx-siA1ZOg&r=Fe4FicGBU_20P2yihxV-apaNSFb6BSj6AlkptSF2gMk&m=hyCifTgVFRiWqGNv7vZbUzgGYvtZRONGPKU26Y9LEGE&s=CBzxDUHh08AYfRoij2YifRst0RrpYsZYTMEwnPTp8fc&e=
.
-r
On Tue, May 28, 2019 at 2:25 PM Rodric Rabbah <ro...@gmail.com> wrote:
> I created an issue for this test failure here:
>
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_apache_incubator-2Dopenwhisk-2Dapigateway_issues_342&d=DwIBaQ&c=jf_iaSHvJObTbx-siA1ZOg&r=Fe4FicGBU_20P2yihxV-apaNSFb6BSj6AlkptSF2gMk&m=hyCifTgVFRiWqGNv7vZbUzgGYvtZRONGPKU26Y9LEGE&s=rfwCCP1WUkA52IPE1uXmkVa8O5nrZ7nCBYV5VDBr_S4&e=
>
> Which implicates this line:
> tenants.lua:130: attempt to compare nil with number
>
>
>
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_apache_incubator-2Dopenwhisk-2Dapigateway_blob_1f46de9d2f7a9420e48c533af7a4b1b6118aa130_scripts_lua_management_lib_tenants.lua-23L129-2DL130&d=DwIBaQ&c=jf_iaSHvJObTbx-siA1ZOg&r=Fe4FicGBU_20P2yihxV-apaNSFb6BSj6AlkptSF2gMk&m=hyCifTgVFRiWqGNv7vZbUzgGYvtZRONGPKU26Y9LEGE&s=1PJ3EYbcLKghDeG5VG_5roSrQ8fvmLIwdcnHHnzbq8Q&e=
>
> Lars - maybe you can spot the issue from that? I suspect guarding against
> nil is sufficient.
>
> -r
>
>
> On Tue, May 28, 2019 at 1:44 PM Lars Andersson <
> lars.storfiskarn.andersson@gmail.com> wrote:
>
>> If that PR is the cause, feel free to revert it as I have no time
>> available
>> right now to have a look.
>>
>> / Lars
>>
>> On Tue, 28 May 2019, 19:35 Matt Hamann <matthew.hamann@gmail.com wrote:
>>
>> > Looks like the failures began directly after the merging of PR #335.
>> Rodric
>> > and Lars were the contributors of that PR.
>> >
>> > I would elect to either revert the PR or ask one of the aforementioned
>> to
>> > investigate the failure and submit a fix. The cause of the failure is
>> not
>> > obvious from the build output.
>> >
>> > ------------------------------------------------
>> > -Matt
>> > matthew.hamann@gmail.com
>> >
>> >
>> > On Tue, May 28, 2019 at 12:03 PM David P Grove <gr...@us.ibm.com>
>> wrote:
>> >
>> > >
>> > > Hi,
>> > >
>> > > It's been 20 days since the daily test run for apigateway
has
>> > > actually succeeded. Spot-checking a few of the runs, it seems like
a
>> > > consistent failure of one test case:
>> > >
>> > > apigw.healthtests.ApiGwRestEndToEndTests > Wsk api should create an
>> API
>> > and
>> > > successfully invoke that API FAILED
>> > > org.scalatest.exceptions.TestFailedException: 502 was not equal
to
>> > 200
>> > >
>> > > Does anyone have any insight into why this test is now
>> failing?
>> > >
>> > > thanks,
>> > >
>> > > --dave
>> > >
>> >
>>
>
Re: apigateway TravisCI test failure on master for almost 3 weeks
Posted by Rodric Rabbah <ro...@gmail.com>.
I posted an explanation to the issue and opened a suggested patch
https://github.com/apache/incubator-openwhisk-apigateway/pull/343.
-r
On Tue, May 28, 2019 at 2:25 PM Rodric Rabbah <ro...@gmail.com> wrote:
> I created an issue for this test failure here:
> https://github.com/apache/incubator-openwhisk-apigateway/issues/342
>
> Which implicates this line:
> tenants.lua:130: attempt to compare nil with number
>
>
> https://github.com/apache/incubator-openwhisk-apigateway/blob/1f46de9d2f7a9420e48c533af7a4b1b6118aa130/scripts/lua/management/lib/tenants.lua#L129-L130
>
> Lars - maybe you can spot the issue from that? I suspect guarding against
> nil is sufficient.
>
> -r
>
>
> On Tue, May 28, 2019 at 1:44 PM Lars Andersson <
> lars.storfiskarn.andersson@gmail.com> wrote:
>
>> If that PR is the cause, feel free to revert it as I have no time
>> available
>> right now to have a look.
>>
>> / Lars
>>
>> On Tue, 28 May 2019, 19:35 Matt Hamann <matthew.hamann@gmail.com wrote:
>>
>> > Looks like the failures began directly after the merging of PR #335.
>> Rodric
>> > and Lars were the contributors of that PR.
>> >
>> > I would elect to either revert the PR or ask one of the aforementioned
>> to
>> > investigate the failure and submit a fix. The cause of the failure is
>> not
>> > obvious from the build output.
>> >
>> > ------------------------------------------------
>> > -Matt
>> > matthew.hamann@gmail.com
>> >
>> >
>> > On Tue, May 28, 2019 at 12:03 PM David P Grove <gr...@us.ibm.com>
>> wrote:
>> >
>> > >
>> > > Hi,
>> > >
>> > > It's been 20 days since the daily test run for apigateway has
>> > > actually succeeded. Spot-checking a few of the runs, it seems like a
>> > > consistent failure of one test case:
>> > >
>> > > apigw.healthtests.ApiGwRestEndToEndTests > Wsk api should create an
>> API
>> > and
>> > > successfully invoke that API FAILED
>> > > org.scalatest.exceptions.TestFailedException: 502 was not equal to
>> > 200
>> > >
>> > > Does anyone have any insight into why this test is now
>> failing?
>> > >
>> > > thanks,
>> > >
>> > > --dave
>> > >
>> >
>>
>
Re: apigateway TravisCI test failure on master for almost 3 weeks
Posted by Rodric Rabbah <ro...@gmail.com>.
I created an issue for this test failure here:
https://github.com/apache/incubator-openwhisk-apigateway/issues/342
Which implicates this line:
tenants.lua:130: attempt to compare nil with number
https://github.com/apache/incubator-openwhisk-apigateway/blob/1f46de9d2f7a9420e48c533af7a4b1b6118aa130/scripts/lua/management/lib/tenants.lua#L129-L130
Lars - maybe you can spot the issue from that? I suspect guarding against
nil is sufficient.
-r
On Tue, May 28, 2019 at 1:44 PM Lars Andersson <
lars.storfiskarn.andersson@gmail.com> wrote:
> If that PR is the cause, feel free to revert it as I have no time available
> right now to have a look.
>
> / Lars
>
> On Tue, 28 May 2019, 19:35 Matt Hamann <matthew.hamann@gmail.com wrote:
>
> > Looks like the failures began directly after the merging of PR #335.
> Rodric
> > and Lars were the contributors of that PR.
> >
> > I would elect to either revert the PR or ask one of the aforementioned to
> > investigate the failure and submit a fix. The cause of the failure is not
> > obvious from the build output.
> >
> > ------------------------------------------------
> > -Matt
> > matthew.hamann@gmail.com
> >
> >
> > On Tue, May 28, 2019 at 12:03 PM David P Grove <gr...@us.ibm.com>
> wrote:
> >
> > >
> > > Hi,
> > >
> > > It's been 20 days since the daily test run for apigateway has
> > > actually succeeded. Spot-checking a few of the runs, it seems like a
> > > consistent failure of one test case:
> > >
> > > apigw.healthtests.ApiGwRestEndToEndTests > Wsk api should create an API
> > and
> > > successfully invoke that API FAILED
> > > org.scalatest.exceptions.TestFailedException: 502 was not equal to
> > 200
> > >
> > > Does anyone have any insight into why this test is now failing?
> > >
> > > thanks,
> > >
> > > --dave
> > >
> >
>
Re: apigateway TravisCI test failure on master for almost 3 weeks
Posted by Lars Andersson <la...@gmail.com>.
If that PR is the cause, feel free to revert it as I have no time available
right now to have a look.
/ Lars
On Tue, 28 May 2019, 19:35 Matt Hamann <matthew.hamann@gmail.com wrote:
> Looks like the failures began directly after the merging of PR #335. Rodric
> and Lars were the contributors of that PR.
>
> I would elect to either revert the PR or ask one of the aforementioned to
> investigate the failure and submit a fix. The cause of the failure is not
> obvious from the build output.
>
> ------------------------------------------------
> -Matt
> matthew.hamann@gmail.com
>
>
> On Tue, May 28, 2019 at 12:03 PM David P Grove <gr...@us.ibm.com> wrote:
>
> >
> > Hi,
> >
> > It's been 20 days since the daily test run for apigateway has
> > actually succeeded. Spot-checking a few of the runs, it seems like a
> > consistent failure of one test case:
> >
> > apigw.healthtests.ApiGwRestEndToEndTests > Wsk api should create an API
> and
> > successfully invoke that API FAILED
> > org.scalatest.exceptions.TestFailedException: 502 was not equal to
> 200
> >
> > Does anyone have any insight into why this test is now failing?
> >
> > thanks,
> >
> > --dave
> >
>
Re: apigateway TravisCI test failure on master for almost 3 weeks
Posted by Matt Hamann <ma...@gmail.com>.
Looks like the failures began directly after the merging of PR #335. Rodric
and Lars were the contributors of that PR.
I would elect to either revert the PR or ask one of the aforementioned to
investigate the failure and submit a fix. The cause of the failure is not
obvious from the build output.
------------------------------------------------
-Matt
matthew.hamann@gmail.com
On Tue, May 28, 2019 at 12:03 PM David P Grove <gr...@us.ibm.com> wrote:
>
> Hi,
>
> It's been 20 days since the daily test run for apigateway has
> actually succeeded. Spot-checking a few of the runs, it seems like a
> consistent failure of one test case:
>
> apigw.healthtests.ApiGwRestEndToEndTests > Wsk api should create an API and
> successfully invoke that API FAILED
> org.scalatest.exceptions.TestFailedException: 502 was not equal to 200
>
> Does anyone have any insight into why this test is now failing?
>
> thanks,
>
> --dave
>