[ovirt-devel] Re: [VDSM] Travis builds still fail on .coverage rename

2018-07-08 Thread Edward Haas
On Sun, Jul 8, 2018 at 1:42 PM, Nir Soffer  wrote:

> On Sat, Jul 7, 2018 at 9:11 AM Edward Haas  wrote:
>
>> On Sat, Jul 7, 2018 at 9:02 AM, Edward Haas  wrote:
>>
>>>
>>>
>>> On Fri, Jul 6, 2018 at 9:16 PM, Nir Soffer  wrote:
>>>
 On Fri, Jul 6, 2018 at 7:05 PM Edward Haas  wrote:

>
>
> On 6 Jul 2018, at 18:41, Nir Soffer  wrote:
>
>
>
> On Fri, 6 Jul 2018, 18:25 Edward Haas,  wrote:
>
>>
>>
>> On 6 Jul 2018, at 14:35, Nir Soffer  wrote:
>>
>> On Fri, Jul 6, 2018 at 1:12 PM Edward Haas  wrote:
>>
>>> I do not know if it is relevant or not, but the tests that travis
>>> runs for master are taken from the 4.2 branch.
>>> OVS tests are now running using pytest.
>>>
>>
>> What do you mean by "taken from 4.2 branch"?
>>
>>
>> I mean that the branch checked out is 4.2 and not master. It even
>> says so on the console output.
>>
>
> Can you share the url of that build?
>
>
> I just clicked the icon on the vdsm repo: https://travis-ci.org/
> oVirt/vdsm
>

 This is indeed 4.2 build. Any commit in github is tested in travis.
 We would like to fix also the 4.2 builds, but first we need to fix
 master builds.

 You can see here that master build fail:
 https://travis-ci.org/oVirt/vdsm/builds

 Since we added gbd and python-debuginfo:
 https://travis-ci.org/oVirt/vdsm/builds/400644077

 - centos build fail (network-py27)
   https://travis-ci.org/oVirt/vdsm/jobs/400644079

 - fedora 28 build pass
   https://travis-ci.org/oVirt/vdsm/jobs/400644081

 - fedora rawhide fail because we cannot rebuild the image,
   python-libblokdev is missing in rawhide.
   https://travis-ci.org/oVirt/vdsm/jobs/400644083
   See https://lists.ovirt.org/archives/list/devel@ovirt.org/thread/
 CDNETITY5RYOCQBIQQF2NUF6RAHGJRPW/


  I don't know anything about these tests, but this failure looks like:

 1. first test has a timeout
 2. first test cleanup did not run because the cleanup code is not
 correct
 3. second test fail because the first test did not clean up

 This looks like real issue in the code.

>>>
>>> This is the same problem we had on oVirt CI, there are linux bridges on
>>> the node.
>>> I have posted a patch to fail earlier and how the real problem:
>>> https://gerrit.ovirt.org/#/c/92867/
>>> The travis-ci run for it is here: https://travis-ci.org/EdDev/
>>> vdsm/jobs/401143906
>>> This is the problem:
>>> cmdutils.py 151 DEBUG /usr/share/openvswitch/scripts/ovs-ctl
>>> --system-id=random start (cwd None)
>>> cmdutils.py 159 DEBUG FAILED:  = 'rmmod: ERROR: Module bridge is in
>>> use by: br_netfilter\n';  = 1
>>>
>>
> Who is using rmmod?
>

The ovs service is trying to load the ovs kmod, for doing so it needs to
take down the bridge one and reload it after the ovs one.


>
>> Any idea who is creating the "br_netfilter" bridge? I guess this is
>>> travis-ci related.
>>>
>>
> Why do we care about br_netfilter? do we require a system without any
> bridge?
>

Yes, in case ovs kmod has not been loaded in advance.


>
>
>> Actually, this may be Docker or some other package that is
>> installed/setup on it.
>> How can I run the docker with the tests locally to debug this?
>>
>
> Run this in vdsm root directory (copied from .travis.yml):
>
> export DOCKER_IMAGE=ovirtorg/vdsm-test-centos
>
> docker pull $DOCKER_IMAGE
>
> docker run \
> --env TRAVIS_CI=1 \
> --privileged \
> --rm \
> -it \
> -v `pwd`:/vdsm:Z \
> $DOCKER_IMAGE \
> bash -c "cd /vdsm && ./autogen.sh --system && make && make --jobs=2
> check"
>
> Since this is privileged container, you probably want to run this inside a
> vm.
>

OK, will try. But I think the kmod is up to the machine the docker runs in,
so in this case it is the travis slave.


>
>> We run "make check" both in travis (.travis.yml) and ovirt ci
>> (automation/check-patch.sh)
>>
>>
>>>
>>>
>>> On Fri, Jul 6, 2018 at 12:51 AM, Nir Soffer 
>>> wrote:
>>>


 On Thu, Jul 5, 2018 at 10:55 PM Nir Soffer 
 wrote:

> On Thu, Jul 5, 2018 at 5:53 PM Nir Soffer 
> wrote:
>
>> On Thu, Jul 5, 2018 at 5:43 PM Dan Kenigsberg 
>> wrote:
>>
>>> On Thu, Jul 5, 2018 at 2:52 AM, Nir Soffer 
>>> wrote:
>>> > On Wed, Jul 4, 2018 at 1:00 PM Dan Kenigsberg <
>>> dan...@redhat.com> wrote:
>>> >>
>>> >> On Wed, Jul 4, 2018 at 12:48 PM, Nir Soffer <
>>> nsof...@redhat.com> wrote:
>>> >> > Dan, travis build still fail when renaming coverage file
>>> even after
>>> >> > your last patch.
>>> >> >
>>> >> >
>>> >> >
>>> >> > ...SS.
>>> SS...

[ovirt-devel] Re: [VDSM] Travis builds still fail on .coverage rename

2018-07-08 Thread Nir Soffer
On Sat, Jul 7, 2018 at 9:11 AM Edward Haas  wrote:

> On Sat, Jul 7, 2018 at 9:02 AM, Edward Haas  wrote:
>
>>
>>
>> On Fri, Jul 6, 2018 at 9:16 PM, Nir Soffer  wrote:
>>
>>> On Fri, Jul 6, 2018 at 7:05 PM Edward Haas  wrote:
>>>


 On 6 Jul 2018, at 18:41, Nir Soffer  wrote:



 On Fri, 6 Jul 2018, 18:25 Edward Haas,  wrote:

>
>
> On 6 Jul 2018, at 14:35, Nir Soffer  wrote:
>
> On Fri, Jul 6, 2018 at 1:12 PM Edward Haas  wrote:
>
>> I do not know if it is relevant or not, but the tests that travis
>> runs for master are taken from the 4.2 branch.
>> OVS tests are now running using pytest.
>>
>
> What do you mean by "taken from 4.2 branch"?
>
>
> I mean that the branch checked out is 4.2 and not master. It even says
> so on the console output.
>

 Can you share the url of that build?


 I just clicked the icon on the vdsm repo:
 https://travis-ci.org/oVirt/vdsm

>>>
>>> This is indeed 4.2 build. Any commit in github is tested in travis.
>>> We would like to fix also the 4.2 builds, but first we need to fix
>>> master builds.
>>>
>>> You can see here that master build fail:
>>> https://travis-ci.org/oVirt/vdsm/builds
>>>
>>> Since we added gbd and python-debuginfo:
>>> https://travis-ci.org/oVirt/vdsm/builds/400644077
>>>
>>> - centos build fail (network-py27)
>>>   https://travis-ci.org/oVirt/vdsm/jobs/400644079
>>>
>>> - fedora 28 build pass
>>>   https://travis-ci.org/oVirt/vdsm/jobs/400644081
>>>
>>> - fedora rawhide fail because we cannot rebuild the image,
>>>   python-libblokdev is missing in rawhide.
>>>   https://travis-ci.org/oVirt/vdsm/jobs/400644083
>>>   See
>>> https://lists.ovirt.org/archives/list/devel@ovirt.org/thread/CDNETITY5RYOCQBIQQF2NUF6RAHGJRPW/
>>>
>>>
>>>  I don't know anything about these tests, but this failure looks like:
>>>
>>> 1. first test has a timeout
>>> 2. first test cleanup did not run because the cleanup code is not correct
>>> 3. second test fail because the first test did not clean up
>>>
>>> This looks like real issue in the code.
>>>
>>
>> This is the same problem we had on oVirt CI, there are linux bridges on
>> the node.
>> I have posted a patch to fail earlier and how the real problem:
>> https://gerrit.ovirt.org/#/c/92867/
>> The travis-ci run for it is here:
>> https://travis-ci.org/EdDev/vdsm/jobs/401143906
>> This is the problem:
>> cmdutils.py 151 DEBUG /usr/share/openvswitch/scripts/ovs-ctl
>> --system-id=random start (cwd None)
>> cmdutils.py 159 DEBUG FAILED:  = 'rmmod: ERROR: Module bridge is in
>> use by: br_netfilter\n';  = 1
>>
>
Who is using rmmod?


> Any idea who is creating the "br_netfilter" bridge? I guess this is
>> travis-ci related.
>>
>
Why do we care about br_netfilter? do we require a system without any
bridge?


> Actually, this may be Docker or some other package that is installed/setup
> on it.
> How can I run the docker with the tests locally to debug this?
>

Run this in vdsm root directory (copied from .travis.yml):

export DOCKER_IMAGE=ovirtorg/vdsm-test-centos

docker pull $DOCKER_IMAGE

docker run \
--env TRAVIS_CI=1 \
--privileged \
--rm \
-it \
-v `pwd`:/vdsm:Z \
$DOCKER_IMAGE \
bash -c "cd /vdsm && ./autogen.sh --system && make && make --jobs=2
check"

Since this is privileged container, you probably want to run this inside a
vm.


> We run "make check" both in travis (.travis.yml) and ovirt ci
> (automation/check-patch.sh)
>
>
>>
>>
>> On Fri, Jul 6, 2018 at 12:51 AM, Nir Soffer 
>> wrote:
>>
>>>
>>>
>>> On Thu, Jul 5, 2018 at 10:55 PM Nir Soffer 
>>> wrote:
>>>
 On Thu, Jul 5, 2018 at 5:53 PM Nir Soffer 
 wrote:

> On Thu, Jul 5, 2018 at 5:43 PM Dan Kenigsberg 
> wrote:
>
>> On Thu, Jul 5, 2018 at 2:52 AM, Nir Soffer 
>> wrote:
>> > On Wed, Jul 4, 2018 at 1:00 PM Dan Kenigsberg <
>> dan...@redhat.com> wrote:
>> >>
>> >> On Wed, Jul 4, 2018 at 12:48 PM, Nir Soffer <
>> nsof...@redhat.com> wrote:
>> >> > Dan, travis build still fail when renaming coverage file
>> even after
>> >> > your last patch.
>> >> >
>> >> >
>> >> >
>> >> >
>> ...SS.SS.SS..S.SSSS.SSS...SSS...S.S.SSSS..S.SS...

[ovirt-devel] Re: [VDSM] Travis builds still fail on .coverage rename

2018-07-07 Thread Daniel Belenky
On Thu, Jul 5, 2018 at 17:56 Nir Soffer  wrote:

> On Thu, Jul 5, 2018 at 5:43 PM Dan Kenigsberg  wrote:
>
>> On Thu, Jul 5, 2018 at 2:52 AM, Nir Soffer  wrote:
>> > On Wed, Jul 4, 2018 at 1:00 PM Dan Kenigsberg 
>> wrote:
>> >>
>> >> On Wed, Jul 4, 2018 at 12:48 PM, Nir Soffer 
>> wrote:
>> >> > Dan, travis build still fail when renaming coverage file even after
>> >> > your last patch.
>> >> >
>> >> >
>> >> >
>> >> >
>> ...SS.SS.SS..S.SSSS.SSS...SSS...S.S.SSSS..S.SS..
>> >> >
>> --
>> >> > Ran 1267 tests in 99.239s
>> >> > OK (SKIP=63)
>> >> > [ -n "$NOSE_WITH_COVERAGE" ] && mv .coverage .coverage-nose-py2
>> >> > make[1]: *** [check] Error 1
>> >> > make[1]: Leaving directory `/vdsm/tests'
>> >> > ERROR: InvocationError: '/usr/bin/make -C tests check'
>> >> >
>> >> > https://travis-ci.org/oVirt/vdsm/jobs/399932012
>> >> >
>> >> > Do you have any idea what is wrong there?
>> >> >
>> >> > Why we don't have any error message from the failed command?
>> >>
>> >> No idea, nothing pops to mind.
>> >> We can revert to the sillier [ -f .coverage ] condition instead of
>> >> understanding (yeah, this feels dirty)
>> >
>> >
>> > Thanks, your patch (https://gerrit.ovirt.org/#/c/92813/) fixed this
>> > failure.
>> >
>> > Now we have failures for the pywatch_test, and some network
>> > tests. Can someone from network look at this?
>> > https://travis-ci.org/nirs/vdsm/builds/400204807
>>
>> https://travis-ci.org/nirs/vdsm/jobs/400204808 shows
>>
>>   ConfigNetworkError: (21, 'Executing commands failed:
>> ovs-vsctl: cannot create a bridge named vdsmbr_test because a bridge
>> named vdsmbr_test already exists')
>>
>> which I thought was limited to dirty ovirt-ci jenkins slaves. Any idea
>> why it shows here?
>>
>
> Maybe one failed test leave dirty host to the next test?
>
>
>> py-watch seems to be failing due to missing gdb on the travis image
>
>
>> cmdutils.py151 DEBUG./py-watch 0.1 sleep 10 (cwd None)
>> cmdutils.py159 DEBUGFAILED:  = 'Traceback
>> (most recent call last):\n  File "./py-watch", line 60, in \n
>>   dump_trace(watched_proc)\n  File "./py-watch", line 32, in
>> dump_trace\n\'thread apply all py-bt\'])\n  File
>> "/usr/lib64/python2.7/site-packages/subprocess32.py", line 575, in
>> call\np = Popen(*popenargs, **kwargs)\n  File
>> "/usr/lib64/python2.7/site-packages/subprocess32.py", line 822, in
>> __init__\nrestore_signals, start_new_session)\n  File
>> "/usr/lib64/python2.7/site-packages/subprocess32.py", line 1567, in
>> _execute_child\nraise child_exception_type(errno_num,
>> err_msg)\nOSError: [Errno 2] No such file or directory: \'gdb\'\n';
>>  = 1
>>
>
> Cool, easy fix.
>
>
>> Nir, could you remind me what is "ERROR: InterpreterNotFound:
>> python3.6" and how can we avoid it? it keeps distracting during
>> debugging test failures.
>>
>
> We can avoid it in travis using env matrix.
>
> Currently we run "make check" which run all the the tox envs
> (e.g. storage-py27,storage-py36) regardless of the build type. This is good
> for manual usage when you don't know which python version is available
> on a developer machine. For example if I have python 3.7 installed, maybe
> I like to test.
>
> We can change this so we will test only the *-py27 on centos, and both
> *-py27 and *-py36 on Fedora.
>
> We can do the same in ovirt CI but it will be harder, we don't have a
> declerative
> way to configure this.
>

This behavior could also be achieved in STDCI V1 relatively easily by
having different scripts and configurations for CentOS and Fedora. I’d be
glad to help to setup such configuration if needed.

BTW, we actually do support a YAML configuration for exactly such cases
(STDCI V2). We are currently proactively migrating smaller oVirt projects
(left VDSM and engine to be last).


> Nir
> ___
> Devel mailing list -- devel@ovirt.org
> To unsubscribe send an email to devel-le...@ovirt.org
> Privacy Statement: https://www.ovirt.org/site/privacy-policy/
> oVirt Code of Conduct:
> https://www.ovirt.org/community/about/community-guidelines/
> List Archives:
> https://lists.ovirt.org/archives/list/devel@ovirt.org/message/UKVQVCL2M6NE5VHZQMIPNYNSREIJOQUQ/
>
-- 
Daniel Belenky
RHV DevOps
_

[ovirt-devel] Re: [VDSM] Travis builds still fail on .coverage rename

2018-07-07 Thread Edward Haas


> On 6 Jul 2018, at 18:41, Nir Soffer  wrote:
> 
> 
> 
>> On Fri, 6 Jul 2018, 18:25 Edward Haas,  wrote:
>> 
>> 
>>> On 6 Jul 2018, at 14:35, Nir Soffer  wrote:
>>> 
 On Fri, Jul 6, 2018 at 1:12 PM Edward Haas  wrote:
 I do not know if it is relevant or not, but the tests that travis runs for 
 master are taken from the 4.2 branch.
 OVS tests are now running using pytest.
>>> 
>>> What do you mean by "taken from 4.2 branch"?
>> 
>> I mean that the branch checked out is 4.2 and not master. It even says so on 
>> the console output.
> 
> 
> Can you share the url of that build?

I just clicked the icon on the vdsm repo: https://travis-ci.org/oVirt/vdsm

> 
>> 
>>> 
>>> We run "make check" both in travis (.travis.yml) and ovirt ci 
>>> (automation/check-patch.sh)
>>>  
 
 
> On Fri, Jul 6, 2018 at 12:51 AM, Nir Soffer  wrote:
> 
> 
>> On Thu, Jul 5, 2018 at 10:55 PM Nir Soffer  wrote:
>>> On Thu, Jul 5, 2018 at 5:53 PM Nir Soffer  wrote:
 On Thu, Jul 5, 2018 at 5:43 PM Dan Kenigsberg  
 wrote:
 On Thu, Jul 5, 2018 at 2:52 AM, Nir Soffer  wrote:
 > On Wed, Jul 4, 2018 at 1:00 PM Dan Kenigsberg  
 > wrote:
 >>
 >> On Wed, Jul 4, 2018 at 12:48 PM, Nir Soffer  
 >> wrote:
 >> > Dan, travis build still fail when renaming coverage file even 
 >> > after
 >> > your last patch.
 >> >
 >> >
 >> >
 >> > ...SS.SS.SS..S.SSSS.SSS...SSS...S.S.SSSS..S.SS..
 >> > --
 >> > Ran 1267 tests in 99.239s
 >> > OK (SKIP=63)
 >> > [ -n "$NOSE_WITH_COVERAGE" ] && mv .coverage .coverage-nose-py2
 >> > make[1]: *** [check] Error 1
 >> > make[1]: Leaving directory `/vdsm/tests'
 >> > ERROR: InvocationError: '/usr/bin/make -C tests check'
 >> >
 >> > https://travis-ci.org/oVirt/vdsm/jobs/399932012
 >> >
 >> > Do you have any idea what is wrong there?
 >> >
 >> > Why we don't have any error message from the failed command?
 >>
 >> No idea, nothing pops to mind.
 >> We can revert to the sillier [ -f .coverage ] condition instead of
 >> understanding (yeah, this feels dirty)
 >
 >
 > Thanks, your patch (https://gerrit.ovirt.org/#/c/92813/) fixed this
 > failure.
 >
 > Now we have failures for the pywatch_test, and some network
 > tests. Can someone from network look at this?
 > https://travis-ci.org/nirs/vdsm/builds/400204807
 
 https://travis-ci.org/nirs/vdsm/jobs/400204808 shows
 
   ConfigNetworkError: (21, 'Executing commands failed:
 ovs-vsctl: cannot create a bridge named vdsmbr_test because a bridge
 named vdsmbr_test already exists')
 
 which I thought was limited to dirty ovirt-ci jenkins slaves. Any idea
 why it shows here?
>>> 
>>> Maybe one failed test leave dirty host to the next test?
> 
> network tests fail now only on CentOS now.
>  
>>>  
 py-watch seems to be failing due to missing gdb on the travis image
 
 cmdutils.py151 DEBUG./py-watch 0.1 sleep 10 (cwd 
 None)
 cmdutils.py159 DEBUGFAILED:  = 'Traceback
 (most recent call last):\n  File "./py-watch", line 60, in \n
   dump_trace(watched_proc)\n  File "./py-watch", line 32, in
 dump_trace\n\'thread apply all py-bt\'])\n  File
 "/usr/lib64/python2.7/site-packages/subprocess32.py", line 575, in
 call\np = Popen(*popenargs, **kwargs)\n  File
 "/usr/lib64/python2.7/site-packages/subprocess32.py", line 822, in
 __init__\nrestore_signals, start_new_session)\n  File
 "/usr/lib64/python2.7/site-packages/subprocess32.py", line 1567, in
 _execute_child\nraise child_exception_type(errno_num,
 err_msg)\nOSError: [Errno 2] No such file or directory: \'gdb\'\n';
  = 1
>>> 
>>> Cool, easy fix.
>> 
>> Fixed by https://gerrit.ovirt.org/#/c/92846/
> 
> Fedora 28 build is green with this change:
> https://travis-

[ovirt-devel] Re: [VDSM] Travis builds still fail on .coverage rename

2018-07-07 Thread Edward Haas


> On 6 Jul 2018, at 14:35, Nir Soffer  wrote:
> 
>> On Fri, Jul 6, 2018 at 1:12 PM Edward Haas  wrote:
>> I do not know if it is relevant or not, but the tests that travis runs for 
>> master are taken from the 4.2 branch.
>> OVS tests are now running using pytest.
> 
> What do you mean by "taken from 4.2 branch"?

I mean that the branch checked out is 4.2 and not master. It even says so on 
the console output.

> 
> We run "make check" both in travis (.travis.yml) and ovirt ci 
> (automation/check-patch.sh)
>  
>> 
>> 
>>> On Fri, Jul 6, 2018 at 12:51 AM, Nir Soffer  wrote:
>>> 
>>> 
 On Thu, Jul 5, 2018 at 10:55 PM Nir Soffer  wrote:
> On Thu, Jul 5, 2018 at 5:53 PM Nir Soffer  wrote:
>> On Thu, Jul 5, 2018 at 5:43 PM Dan Kenigsberg  wrote:
>> On Thu, Jul 5, 2018 at 2:52 AM, Nir Soffer  wrote:
>> > On Wed, Jul 4, 2018 at 1:00 PM Dan Kenigsberg  
>> > wrote:
>> >>
>> >> On Wed, Jul 4, 2018 at 12:48 PM, Nir Soffer  
>> >> wrote:
>> >> > Dan, travis build still fail when renaming coverage file even after
>> >> > your last patch.
>> >> >
>> >> >
>> >> >
>> >> > ...SS.SS.SS..S.SSSS.SSS...SSS...S.S.SSSS..S.SS..
>> >> > --
>> >> > Ran 1267 tests in 99.239s
>> >> > OK (SKIP=63)
>> >> > [ -n "$NOSE_WITH_COVERAGE" ] && mv .coverage .coverage-nose-py2
>> >> > make[1]: *** [check] Error 1
>> >> > make[1]: Leaving directory `/vdsm/tests'
>> >> > ERROR: InvocationError: '/usr/bin/make -C tests check'
>> >> >
>> >> > https://travis-ci.org/oVirt/vdsm/jobs/399932012
>> >> >
>> >> > Do you have any idea what is wrong there?
>> >> >
>> >> > Why we don't have any error message from the failed command?
>> >>
>> >> No idea, nothing pops to mind.
>> >> We can revert to the sillier [ -f .coverage ] condition instead of
>> >> understanding (yeah, this feels dirty)
>> >
>> >
>> > Thanks, your patch (https://gerrit.ovirt.org/#/c/92813/) fixed this
>> > failure.
>> >
>> > Now we have failures for the pywatch_test, and some network
>> > tests. Can someone from network look at this?
>> > https://travis-ci.org/nirs/vdsm/builds/400204807
>> 
>> https://travis-ci.org/nirs/vdsm/jobs/400204808 shows
>> 
>>   ConfigNetworkError: (21, 'Executing commands failed:
>> ovs-vsctl: cannot create a bridge named vdsmbr_test because a bridge
>> named vdsmbr_test already exists')
>> 
>> which I thought was limited to dirty ovirt-ci jenkins slaves. Any idea
>> why it shows here?
> 
> Maybe one failed test leave dirty host to the next test?
>>> 
>>> network tests fail now only on CentOS now.
>>>  
>  
>> py-watch seems to be failing due to missing gdb on the travis image
>> 
>> cmdutils.py151 DEBUG./py-watch 0.1 sleep 10 (cwd 
>> None)
>> cmdutils.py159 DEBUGFAILED:  = 'Traceback
>> (most recent call last):\n  File "./py-watch", line 60, in \n
>>   dump_trace(watched_proc)\n  File "./py-watch", line 32, in
>> dump_trace\n\'thread apply all py-bt\'])\n  File
>> "/usr/lib64/python2.7/site-packages/subprocess32.py", line 575, in
>> call\np = Popen(*popenargs, **kwargs)\n  File
>> "/usr/lib64/python2.7/site-packages/subprocess32.py", line 822, in
>> __init__\nrestore_signals, start_new_session)\n  File
>> "/usr/lib64/python2.7/site-packages/subprocess32.py", line 1567, in
>> _execute_child\nraise child_exception_type(errno_num,
>> err_msg)\nOSError: [Errno 2] No such file or directory: \'gdb\'\n';
>>  = 1
> 
> Cool, easy fix.
 
 Fixed by https://gerrit.ovirt.org/#/c/92846/
>>> 
>>> Fedora 28 build is green with this change:
>>> https://travis-ci.org/nirs/vdsm/jobs/400549561
>>>  
>>> 
>>> ___ summary 
>>> 
>>>   tests: commands succeeded
>>>   storage-py27: commands succeeded
>>>   storage-py36: commands succeeded
>>>   lib-py27: commands succeeded
>>>   lib-py36: commands succeeded
>>>   network-py27: commands succeeded
>>>   network-py36: commands succeeded
>>>   virt-py27: commands succeeded
>>>   virt-py36: commands succeeded
>>

[ovirt-devel] Re: [VDSM] Travis builds still fail on .coverage rename

2018-07-07 Thread Greg Sheremeta
Hey,

I was in a similar same place with docker networking and webdriver in
Jenkins a few months ago :)

Best way to mimic CI locally is using mock_runner. [I tried on a fresh
machine yesterday and got stuck -- will try again soon, it runs on my old
machine great ]

https://ovirt-jira.atlassian.net/browse/OVIRT-2275
https://www.ovirt.org/blog/2017/01/ovirt-system-tests-to-the-rescue/

The magic incantation on my older machine (for build-artifacts, but the
last flag switches the stage):
cd /my/project
../jenkins/mock_configs/mock_runner.sh --mock-confs-dir
../jenkins/mock_configs fc28:fedora-28-x86_64 --build-only

Best wishes,
Greg

On Sat, Jul 7, 2018 at 2:45 AM Edward Haas  wrote:

>
>
> On Sat, Jul 7, 2018 at 9:02 AM, Edward Haas  wrote:
>
>>
>>
>> On Fri, Jul 6, 2018 at 9:16 PM, Nir Soffer  wrote:
>>
>>> On Fri, Jul 6, 2018 at 7:05 PM Edward Haas  wrote:
>>>


 On 6 Jul 2018, at 18:41, Nir Soffer  wrote:



 On Fri, 6 Jul 2018, 18:25 Edward Haas,  wrote:

>
>
> On 6 Jul 2018, at 14:35, Nir Soffer  wrote:
>
> On Fri, Jul 6, 2018 at 1:12 PM Edward Haas  wrote:
>
>> I do not know if it is relevant or not, but the tests that travis
>> runs for master are taken from the 4.2 branch.
>> OVS tests are now running using pytest.
>>
>
> What do you mean by "taken from 4.2 branch"?
>
>
> I mean that the branch checked out is 4.2 and not master. It even says
> so on the console output.
>

 Can you share the url of that build?


 I just clicked the icon on the vdsm repo:
 https://travis-ci.org/oVirt/vdsm

>>>
>>> This is indeed 4.2 build. Any commit in github is tested in travis.
>>> We would like to fix also the 4.2 builds, but first we need to fix
>>> master builds.
>>>
>>> You can see here that master build fail:
>>> https://travis-ci.org/oVirt/vdsm/builds
>>>
>>> Since we added gbd and python-debuginfo:
>>> https://travis-ci.org/oVirt/vdsm/builds/400644077
>>>
>>> - centos build fail (network-py27)
>>>   https://travis-ci.org/oVirt/vdsm/jobs/400644079
>>>
>>> - fedora 28 build pass
>>>   https://travis-ci.org/oVirt/vdsm/jobs/400644081
>>>
>>> - fedora rawhide fail because we cannot rebuild the image,
>>>   python-libblokdev is missing in rawhide.
>>>   https://travis-ci.org/oVirt/vdsm/jobs/400644083
>>>   See
>>> https://lists.ovirt.org/archives/list/devel@ovirt.org/thread/CDNETITY5RYOCQBIQQF2NUF6RAHGJRPW/
>>>
>>>
>>>  I don't know anything about these tests, but this failure looks like:
>>>
>>> 1. first test has a timeout
>>> 2. first test cleanup did not run because the cleanup code is not correct
>>> 3. second test fail because the first test did not clean up
>>>
>>> This looks like real issue in the code.
>>>
>>
>> This is the same problem we had on oVirt CI, there are linux bridges on
>> the node.
>> I have posted a patch to fail earlier and how the real problem:
>> https://gerrit.ovirt.org/#/c/92867/
>> The travis-ci run for it is here:
>> https://travis-ci.org/EdDev/vdsm/jobs/401143906
>> This is the problem:
>> cmdutils.py 151 DEBUG /usr/share/openvswitch/scripts/ovs-ctl
>> --system-id=random start (cwd None)
>> cmdutils.py 159 DEBUG FAILED:  = 'rmmod: ERROR: Module bridge is in
>> use by: br_netfilter\n';  = 1
>>
>> Any idea who is creating the "br_netfilter" bridge? I guess this is
>> travis-ci related.
>>
>
> Actually, this may be Docker or some other package that is installed/setup
> on it.
> How can I run the docker with the tests locally to debug this?
>
>
>>
>>
> Thanks,
>> Edy.
>>
>>
>>>

>
> We run "make check" both in travis (.travis.yml) and ovirt ci
> (automation/check-patch.sh)
>
>
>>
>>
>> On Fri, Jul 6, 2018 at 12:51 AM, Nir Soffer 
>> wrote:
>>
>>>
>>>
>>> On Thu, Jul 5, 2018 at 10:55 PM Nir Soffer 
>>> wrote:
>>>
 On Thu, Jul 5, 2018 at 5:53 PM Nir Soffer 
 wrote:

> On Thu, Jul 5, 2018 at 5:43 PM Dan Kenigsberg 
> wrote:
>
>> On Thu, Jul 5, 2018 at 2:52 AM, Nir Soffer 
>> wrote:
>> > On Wed, Jul 4, 2018 at 1:00 PM Dan Kenigsberg <
>> dan...@redhat.com> wrote:
>> >>
>> >> On Wed, Jul 4, 2018 at 12:48 PM, Nir Soffer <
>> nsof...@redhat.com> wrote:
>> >> > Dan, travis build still fail when renaming coverage file
>> even after
>> >> > your last patch.
>> >> >
>> >> >
>> >> >
>> >> >
>> ...SS.SS.SS..S.SSSS.SSS...SSS...S.S.S...

[ovirt-devel] Re: [VDSM] Travis builds still fail on .coverage rename

2018-07-06 Thread Edward Haas
On Sat, Jul 7, 2018 at 9:02 AM, Edward Haas  wrote:

>
>
> On Fri, Jul 6, 2018 at 9:16 PM, Nir Soffer  wrote:
>
>> On Fri, Jul 6, 2018 at 7:05 PM Edward Haas  wrote:
>>
>>>
>>>
>>> On 6 Jul 2018, at 18:41, Nir Soffer  wrote:
>>>
>>>
>>>
>>> On Fri, 6 Jul 2018, 18:25 Edward Haas,  wrote:
>>>


 On 6 Jul 2018, at 14:35, Nir Soffer  wrote:

 On Fri, Jul 6, 2018 at 1:12 PM Edward Haas  wrote:

> I do not know if it is relevant or not, but the tests that travis runs
> for master are taken from the 4.2 branch.
> OVS tests are now running using pytest.
>

 What do you mean by "taken from 4.2 branch"?


 I mean that the branch checked out is 4.2 and not master. It even says
 so on the console output.

>>>
>>> Can you share the url of that build?
>>>
>>>
>>> I just clicked the icon on the vdsm repo: https://travis-ci.org/oV
>>> irt/vdsm
>>>
>>
>> This is indeed 4.2 build. Any commit in github is tested in travis.
>> We would like to fix also the 4.2 builds, but first we need to fix master
>> builds.
>>
>> You can see here that master build fail:
>> https://travis-ci.org/oVirt/vdsm/builds
>>
>> Since we added gbd and python-debuginfo:
>> https://travis-ci.org/oVirt/vdsm/builds/400644077
>>
>> - centos build fail (network-py27)
>>   https://travis-ci.org/oVirt/vdsm/jobs/400644079
>>
>> - fedora 28 build pass
>>   https://travis-ci.org/oVirt/vdsm/jobs/400644081
>>
>> - fedora rawhide fail because we cannot rebuild the image,
>>   python-libblokdev is missing in rawhide.
>>   https://travis-ci.org/oVirt/vdsm/jobs/400644083
>>   See https://lists.ovirt.org/archives/list/devel@ovirt.org/th
>> read/CDNETITY5RYOCQBIQQF2NUF6RAHGJRPW/
>>
>>
>>  I don't know anything about these tests, but this failure looks like:
>>
>> 1. first test has a timeout
>> 2. first test cleanup did not run because the cleanup code is not correct
>> 3. second test fail because the first test did not clean up
>>
>> This looks like real issue in the code.
>>
>
> This is the same problem we had on oVirt CI, there are linux bridges on
> the node.
> I have posted a patch to fail earlier and how the real problem:
> https://gerrit.ovirt.org/#/c/92867/
> The travis-ci run for it is here: https://travis-ci.org/EdDev/
> vdsm/jobs/401143906
> This is the problem:
> cmdutils.py 151 DEBUG /usr/share/openvswitch/scripts/ovs-ctl
> --system-id=random start (cwd None)
> cmdutils.py 159 DEBUG FAILED:  = 'rmmod: ERROR: Module bridge is in
> use by: br_netfilter\n';  = 1
>
> Any idea who is creating the "br_netfilter" bridge? I guess this is
> travis-ci related.
>

Actually, this may be Docker or some other package that is installed/setup
on it.
How can I run the docker with the tests locally to debug this?


>
>
Thanks,
> Edy.
>
>
>>
>>>

 We run "make check" both in travis (.travis.yml) and ovirt ci
 (automation/check-patch.sh)


>
>
> On Fri, Jul 6, 2018 at 12:51 AM, Nir Soffer 
> wrote:
>
>>
>>
>> On Thu, Jul 5, 2018 at 10:55 PM Nir Soffer 
>> wrote:
>>
>>> On Thu, Jul 5, 2018 at 5:53 PM Nir Soffer 
>>> wrote:
>>>
 On Thu, Jul 5, 2018 at 5:43 PM Dan Kenigsberg 
 wrote:

> On Thu, Jul 5, 2018 at 2:52 AM, Nir Soffer 
> wrote:
> > On Wed, Jul 4, 2018 at 1:00 PM Dan Kenigsberg 
> wrote:
> >>
> >> On Wed, Jul 4, 2018 at 12:48 PM, Nir Soffer 
> wrote:
> >> > Dan, travis build still fail when renaming coverage file even
> after
> >> > your last patch.
> >> >
> >> >
> >> >
> >> > ...SS.SS
> 
> 
> .SS.
> .S.SS...
> .SS.SS..
> ..S...SSS...S.S.
> S...
> .SSS..S.SS..
> 
> 
> 
> >> > 
> --
> >> > Ran 1267 tests in 99.239s
> >> > OK (SKIP=63)
> >> > [ -n "$NOSE_WITH_COVERAGE" ] && mv .coverage
> .coverage-nose-py2
> >> > make[1]: *** [check] Error 1
> >> > make[1]: Leaving directory `/vdsm/tests'
> >> > ERROR: InvocationError: '/usr/bin/make -C tests check'
> >> >
> >> > https://travis-ci.org/oVirt/vdsm/jobs/399932012
> >> >

[ovirt-devel] Re: [VDSM] Travis builds still fail on .coverage rename

2018-07-06 Thread Edward Haas
On Fri, Jul 6, 2018 at 9:16 PM, Nir Soffer  wrote:

> On Fri, Jul 6, 2018 at 7:05 PM Edward Haas  wrote:
>
>>
>>
>> On 6 Jul 2018, at 18:41, Nir Soffer  wrote:
>>
>>
>>
>> On Fri, 6 Jul 2018, 18:25 Edward Haas,  wrote:
>>
>>>
>>>
>>> On 6 Jul 2018, at 14:35, Nir Soffer  wrote:
>>>
>>> On Fri, Jul 6, 2018 at 1:12 PM Edward Haas  wrote:
>>>
 I do not know if it is relevant or not, but the tests that travis runs
 for master are taken from the 4.2 branch.
 OVS tests are now running using pytest.

>>>
>>> What do you mean by "taken from 4.2 branch"?
>>>
>>>
>>> I mean that the branch checked out is 4.2 and not master. It even says
>>> so on the console output.
>>>
>>
>> Can you share the url of that build?
>>
>>
>> I just clicked the icon on the vdsm repo: https://travis-ci.org/
>> oVirt/vdsm
>>
>
> This is indeed 4.2 build. Any commit in github is tested in travis.
> We would like to fix also the 4.2 builds, but first we need to fix master
> builds.
>
> You can see here that master build fail:
> https://travis-ci.org/oVirt/vdsm/builds
>
> Since we added gbd and python-debuginfo:
> https://travis-ci.org/oVirt/vdsm/builds/400644077
>
> - centos build fail (network-py27)
>   https://travis-ci.org/oVirt/vdsm/jobs/400644079
>
> - fedora 28 build pass
>   https://travis-ci.org/oVirt/vdsm/jobs/400644081
>
> - fedora rawhide fail because we cannot rebuild the image,
>   python-libblokdev is missing in rawhide.
>   https://travis-ci.org/oVirt/vdsm/jobs/400644083
>   See https://lists.ovirt.org/archives/list/devel@ovirt.org/thread/
> CDNETITY5RYOCQBIQQF2NUF6RAHGJRPW/
>
>
>  I don't know anything about these tests, but this failure looks like:
>
> 1. first test has a timeout
> 2. first test cleanup did not run because the cleanup code is not correct
> 3. second test fail because the first test did not clean up
>
> This looks like real issue in the code.
>

This is the same problem we had on oVirt CI, there are linux bridges on the
node.
I have posted a patch to fail earlier and how the real problem:
https://gerrit.ovirt.org/#/c/92867/
The travis-ci run for it is here:
https://travis-ci.org/EdDev/vdsm/jobs/401143906
This is the problem:
cmdutils.py 151 DEBUG /usr/share/openvswitch/scripts/ovs-ctl
--system-id=random start (cwd None)
cmdutils.py 159 DEBUG FAILED:  = 'rmmod: ERROR: Module bridge is in
use by: br_netfilter\n';  = 1

Any idea who is creating the "br_netfilter" bridge? I guess this is
travis-ci related.

Thanks,
Edy.


>
>>
>>>
>>> We run "make check" both in travis (.travis.yml) and ovirt ci
>>> (automation/check-patch.sh)
>>>
>>>


 On Fri, Jul 6, 2018 at 12:51 AM, Nir Soffer  wrote:

>
>
> On Thu, Jul 5, 2018 at 10:55 PM Nir Soffer  wrote:
>
>> On Thu, Jul 5, 2018 at 5:53 PM Nir Soffer  wrote:
>>
>>> On Thu, Jul 5, 2018 at 5:43 PM Dan Kenigsberg 
>>> wrote:
>>>
 On Thu, Jul 5, 2018 at 2:52 AM, Nir Soffer 
 wrote:
 > On Wed, Jul 4, 2018 at 1:00 PM Dan Kenigsberg 
 wrote:
 >>
 >> On Wed, Jul 4, 2018 at 12:48 PM, Nir Soffer 
 wrote:
 >> > Dan, travis build still fail when renaming coverage file even
 after
 >> > your last patch.
 >> >
 >> >
 >> >
 >> > ...SS.SS
 
 
 .SS.
 .S.SS...
 .SS.SS..
 ..S...SSS...S.S.
 S...
 .SSS..S.SS..
 
 
 
 >> > 
 --
 >> > Ran 1267 tests in 99.239s
 >> > OK (SKIP=63)
 >> > [ -n "$NOSE_WITH_COVERAGE" ] && mv .coverage .coverage-nose-py2
 >> > make[1]: *** [check] Error 1
 >> > make[1]: Leaving directory `/vdsm/tests'
 >> > ERROR: InvocationError: '/usr/bin/make -C tests check'
 >> >
 >> > https://travis-ci.org/oVirt/vdsm/jobs/399932012
 >> >
 >> > Do you have any idea what is wrong there?
 >> >
 >> > Why we don't have any error message from the failed command?
 >>
 >> No idea, nothing pops to mind.
 >> We can revert to the sillier [ -f .coverage ] condition instead
 of
 >> understanding (yeah, this feels dirty)
 >
 >
 > Thanks, your patch (https://gerrit.

[ovirt-devel] Re: [VDSM] Travis builds still fail on .coverage rename

2018-07-06 Thread Nir Soffer
On Fri, Jul 6, 2018 at 7:05 PM Edward Haas  wrote:

>
>
> On 6 Jul 2018, at 18:41, Nir Soffer  wrote:
>
>
>
> On Fri, 6 Jul 2018, 18:25 Edward Haas,  wrote:
>
>>
>>
>> On 6 Jul 2018, at 14:35, Nir Soffer  wrote:
>>
>> On Fri, Jul 6, 2018 at 1:12 PM Edward Haas  wrote:
>>
>>> I do not know if it is relevant or not, but the tests that travis runs
>>> for master are taken from the 4.2 branch.
>>> OVS tests are now running using pytest.
>>>
>>
>> What do you mean by "taken from 4.2 branch"?
>>
>>
>> I mean that the branch checked out is 4.2 and not master. It even says so
>> on the console output.
>>
>
> Can you share the url of that build?
>
>
> I just clicked the icon on the vdsm repo: https://travis-ci.org/oVirt/vdsm
>

This is indeed 4.2 build. Any commit in github is tested in travis.
We would like to fix also the 4.2 builds, but first we need to fix master
builds.

You can see here that master build fail:
https://travis-ci.org/oVirt/vdsm/builds

Since we added gbd and python-debuginfo:
https://travis-ci.org/oVirt/vdsm/builds/400644077

- centos build fail (network-py27)
  https://travis-ci.org/oVirt/vdsm/jobs/400644079

- fedora 28 build pass
  https://travis-ci.org/oVirt/vdsm/jobs/400644081

- fedora rawhide fail because we cannot rebuild the image,
  python-libblokdev is missing in rawhide.
  https://travis-ci.org/oVirt/vdsm/jobs/400644083
  See
https://lists.ovirt.org/archives/list/devel@ovirt.org/thread/CDNETITY5RYOCQBIQQF2NUF6RAHGJRPW/


 This is the first failing test:


__ TestOvsApiBase.test_execute_a_transaction ___
self = 
def test_execute_a_transaction(self):
ovsdb = create()
cmd_add_br = ovsdb.add_br(TEST_BRIDGE)
cmd_list_bridge_info = ovsdb.list_bridge_info()
t = ovsdb.transaction()
t.add(cmd_add_br)
t.add(cmd_list_bridge_info)
>   t.commit()
network/integration/ovs/ovs_driver_test.py:51:
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
self = 
def commit(self):
if not self.commands:
return
timeout_option = []
if self.timeout:
timeout_option = ['--timeout=' + str(self.timeout)]
args = []
for command in self.commands:
args += ['--'] + command.cmd
exec_line = [_ovs_vsctl_cmd()] + timeout_option + OUTPUT_FORMAT + args
logging.debug('Executing commands: %s' % ' '.join(exec_line))
rc, out, err = netcmd.exec_sync(exec_line)
if rc != 0:
err = err.splitlines()
if OvsDBConnectionError.is_ovs_db_conn_error(err):
raise OvsDBConnectionError('\n'.join(err))
else:
raise ConfigNetworkError(
ne.ERR_BAD_PARAMS,
>   'Executing commands failed: %s' % '\n'.join(err))
E   ConfigNetworkError: (21, 'Executing commands failed:
2018-07-05T22:24:25Z|2|fatal_signal|WARN|terminating with signal
14 (Alarm clock)')
../lib/vdsm/network/ovs/driver/vsctl.py:78: ConfigNetworkError
-- Captured log call ---
vsctl.py68 DEBUGExecuting commands:
/usr/bin/ovs-vsctl --timeout=5 --oneline --format=json -- add-br
vdsmbr_test -- list Bridge
cmdutils.py151 DEBUG/usr/bin/ovs-vsctl --timeout=5
--oneline --format=json -- add-br vdsmbr_test -- list Bridge (cwd
None)
cmdutils.py159 DEBUGFAILED:  =
'2018-07-05T22:24:25Z|2|fatal_signal|WARN|terminating with signal
14 (Alarm clock)\n';  = -14
 Captured log teardown -
vsctl.py68 DEBUGExecuting commands:
/usr/bin/ovs-vsctl --timeout=5 --oneline --format=json -- list Bridge
cmdutils.py151 DEBUG/usr/bin/ovs-vsctl --timeout=5
--oneline --format=json -- list Bridge (cwd None)
cmdutils.py159 DEBUGSUCCESS:  = '';  = 0

It failed with a timeout.

The second test fail because the bridge already exists:


 TestOvsApiWithSingleRealBridge.test_create_remove_bond 
self = 
def setUp(self):
self.ovsdb = create()
>   self.ovsdb.add_br(TEST_BRIDGE).execute()
network/integration/ovs/ovs_driver_test.py:69:
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
../lib/vdsm/network/ovs/driver/vsctl.py:99: in execute
t.add(self)
../lib/vdsm/network/ovs/driver/__init__.py:46: in __exit__
self.result = self.commit()
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
self = 
def commit(self):
if not self.commands:
return
timeout_option = []
if self.timeout:
timeout_option = ['--timeout=' + str(self.timeout)]
args = []
for command in self.commands:
args += ['--'] + command.cmd
exec_line = [_ovs_vsctl_cmd()] + timeout_option + OUTP

[ovirt-devel] Re: [VDSM] Travis builds still fail on .coverage rename

2018-07-06 Thread Nir Soffer
On Fri, 6 Jul 2018, 18:25 Edward Haas,  wrote:

>
>
> On 6 Jul 2018, at 14:35, Nir Soffer  wrote:
>
> On Fri, Jul 6, 2018 at 1:12 PM Edward Haas  wrote:
>
>> I do not know if it is relevant or not, but the tests that travis runs
>> for master are taken from the 4.2 branch.
>> OVS tests are now running using pytest.
>>
>
> What do you mean by "taken from 4.2 branch"?
>
>
> I mean that the branch checked out is 4.2 and not master. It even says so
> on the console output.
>

Can you share the url of that build?


>
> We run "make check" both in travis (.travis.yml) and ovirt ci
> (automation/check-patch.sh)
>
>
>>
>>
>> On Fri, Jul 6, 2018 at 12:51 AM, Nir Soffer  wrote:
>>
>>>
>>>
>>> On Thu, Jul 5, 2018 at 10:55 PM Nir Soffer  wrote:
>>>
 On Thu, Jul 5, 2018 at 5:53 PM Nir Soffer  wrote:

> On Thu, Jul 5, 2018 at 5:43 PM Dan Kenigsberg 
> wrote:
>
>> On Thu, Jul 5, 2018 at 2:52 AM, Nir Soffer 
>> wrote:
>> > On Wed, Jul 4, 2018 at 1:00 PM Dan Kenigsberg 
>> wrote:
>> >>
>> >> On Wed, Jul 4, 2018 at 12:48 PM, Nir Soffer 
>> wrote:
>> >> > Dan, travis build still fail when renaming coverage file even
>> after
>> >> > your last patch.
>> >> >
>> >> >
>> >> >
>> >> >
>> ...SS.SS.SS..S.SSSS.SSS...SSS...S.S.SSSS..S.SS..
>> >> >
>> --
>> >> > Ran 1267 tests in 99.239s
>> >> > OK (SKIP=63)
>> >> > [ -n "$NOSE_WITH_COVERAGE" ] && mv .coverage .coverage-nose-py2
>> >> > make[1]: *** [check] Error 1
>> >> > make[1]: Leaving directory `/vdsm/tests'
>> >> > ERROR: InvocationError: '/usr/bin/make -C tests check'
>> >> >
>> >> > https://travis-ci.org/oVirt/vdsm/jobs/399932012
>> >> >
>> >> > Do you have any idea what is wrong there?
>> >> >
>> >> > Why we don't have any error message from the failed command?
>> >>
>> >> No idea, nothing pops to mind.
>> >> We can revert to the sillier [ -f .coverage ] condition instead of
>> >> understanding (yeah, this feels dirty)
>> >
>> >
>> > Thanks, your patch (https://gerrit.ovirt.org/#/c/92813/) fixed this
>> > failure.
>> >
>> > Now we have failures for the pywatch_test, and some network
>> > tests. Can someone from network look at this?
>> > https://travis-ci.org/nirs/vdsm/builds/400204807
>>
>> https://travis-ci.org/nirs/vdsm/jobs/400204808 shows
>>
>>   ConfigNetworkError: (21, 'Executing commands failed:
>> ovs-vsctl: cannot create a bridge named vdsmbr_test because a bridge
>> named vdsmbr_test already exists')
>>
>> which I thought was limited to dirty ovirt-ci jenkins slaves. Any idea
>> why it shows here?
>>
>
> Maybe one failed test leave dirty host to the next test?
>

>>> network tests fail now only on CentOS now.
>>>
>>>

>
>> py-watch seems to be failing due to missing gdb on the travis image
>
>
>> cmdutils.py151 DEBUG./py-watch 0.1 sleep 10 (cwd
>> None)
>> cmdutils.py159 DEBUGFAILED:  = 'Traceback
>> (most recent call last):\n  File "./py-watch", line 60, in \n
>>   dump_trace(watched_proc)\n  File "./py-watch", line 32, in
>> dump_trace\n\'thread apply all py-bt\'])\n  File
>> "/usr/lib64/python2.7/site-packages/subprocess32.py", line 575, in
>> call\np = Popen(*popenargs, **kwargs)\n  File
>> "/usr/lib64/python2.7/site-packages/subprocess32.py", line 822, in
>> __init__\nrestore_signals, start_new_session)\n  File
>> "/usr/lib64/python2.7/site-packages/subprocess32.py", line 1567, in
>> _execute_child\nraise child_exception_type(errno_num,
>> err_msg)\nOSError: [Errno 2] No such file or directory: \'gdb\'\n';
>>  = 1
>>
>
> Cool, easy fix.
>

 Fixed by https://gerrit.ovirt.org/#/c/92846/

>>>
>>> Fedora 28 build is green with this change:
>>> https://travis-ci.org/nirs/vdsm/jobs/400549561
>>>
>>>
>>>
>>> ___ summary 
>>> 
>>>   tests: commands succeeded
>>>   storage-py27: commands succeeded
>>>   storage-py36: commands succeeded
>>>   lib-py27: commands succeeded
>>>   lib-py36: comma

[ovirt-devel] Re: [VDSM] Travis builds still fail on .coverage rename

2018-07-06 Thread Nir Soffer
On Fri, Jul 6, 2018 at 1:12 PM Edward Haas  wrote:

> I do not know if it is relevant or not, but the tests that travis runs for
> master are taken from the 4.2 branch.
> OVS tests are now running using pytest.
>

What do you mean by "taken from 4.2 branch"?

We run "make check" both in travis (.travis.yml) and ovirt ci
(automation/check-patch.sh)


>
>
> On Fri, Jul 6, 2018 at 12:51 AM, Nir Soffer  wrote:
>
>>
>>
>> On Thu, Jul 5, 2018 at 10:55 PM Nir Soffer  wrote:
>>
>>> On Thu, Jul 5, 2018 at 5:53 PM Nir Soffer  wrote:
>>>
 On Thu, Jul 5, 2018 at 5:43 PM Dan Kenigsberg 
 wrote:

> On Thu, Jul 5, 2018 at 2:52 AM, Nir Soffer  wrote:
> > On Wed, Jul 4, 2018 at 1:00 PM Dan Kenigsberg 
> wrote:
> >>
> >> On Wed, Jul 4, 2018 at 12:48 PM, Nir Soffer 
> wrote:
> >> > Dan, travis build still fail when renaming coverage file even
> after
> >> > your last patch.
> >> >
> >> >
> >> >
> >> >
> ...SS.SS.SS..S.SSSS.SSS...SSS...S.S.SSSS..S.SS..
> >> >
> --
> >> > Ran 1267 tests in 99.239s
> >> > OK (SKIP=63)
> >> > [ -n "$NOSE_WITH_COVERAGE" ] && mv .coverage .coverage-nose-py2
> >> > make[1]: *** [check] Error 1
> >> > make[1]: Leaving directory `/vdsm/tests'
> >> > ERROR: InvocationError: '/usr/bin/make -C tests check'
> >> >
> >> > https://travis-ci.org/oVirt/vdsm/jobs/399932012
> >> >
> >> > Do you have any idea what is wrong there?
> >> >
> >> > Why we don't have any error message from the failed command?
> >>
> >> No idea, nothing pops to mind.
> >> We can revert to the sillier [ -f .coverage ] condition instead of
> >> understanding (yeah, this feels dirty)
> >
> >
> > Thanks, your patch (https://gerrit.ovirt.org/#/c/92813/) fixed this
> > failure.
> >
> > Now we have failures for the pywatch_test, and some network
> > tests. Can someone from network look at this?
> > https://travis-ci.org/nirs/vdsm/builds/400204807
>
> https://travis-ci.org/nirs/vdsm/jobs/400204808 shows
>
>   ConfigNetworkError: (21, 'Executing commands failed:
> ovs-vsctl: cannot create a bridge named vdsmbr_test because a bridge
> named vdsmbr_test already exists')
>
> which I thought was limited to dirty ovirt-ci jenkins slaves. Any idea
> why it shows here?
>

 Maybe one failed test leave dirty host to the next test?

>>>
>> network tests fail now only on CentOS now.
>>
>>
>>>

> py-watch seems to be failing due to missing gdb on the travis image


> cmdutils.py151 DEBUG./py-watch 0.1 sleep 10 (cwd
> None)
> cmdutils.py159 DEBUGFAILED:  = 'Traceback
> (most recent call last):\n  File "./py-watch", line 60, in \n
>   dump_trace(watched_proc)\n  File "./py-watch", line 32, in
> dump_trace\n\'thread apply all py-bt\'])\n  File
> "/usr/lib64/python2.7/site-packages/subprocess32.py", line 575, in
> call\np = Popen(*popenargs, **kwargs)\n  File
> "/usr/lib64/python2.7/site-packages/subprocess32.py", line 822, in
> __init__\nrestore_signals, start_new_session)\n  File
> "/usr/lib64/python2.7/site-packages/subprocess32.py", line 1567, in
> _execute_child\nraise child_exception_type(errno_num,
> err_msg)\nOSError: [Errno 2] No such file or directory: \'gdb\'\n';
>  = 1
>

 Cool, easy fix.

>>>
>>> Fixed by https://gerrit.ovirt.org/#/c/92846/
>>>
>>
>> Fedora 28 build is green with this change:
>> https://travis-ci.org/nirs/vdsm/jobs/400549561
>>
>>
>>
>> ___ summary 
>> 
>>   tests: commands succeeded
>>   storage-py27: commands succeeded
>>   storage-py36: commands succeeded
>>   lib-py27: commands succeeded
>>   lib-py36: commands succeeded
>>   network-py27: commands succeeded
>>   network-py36: commands succeeded
>>   virt-py27: commands succeeded
>>   virt-py36: commands succeeded
>>   congratulations :)
>>
>>
>>
>>>


> Nir, could you remind me what is "ERROR: InterpreterNotFound:
> python3.6" and how can we avoid it? it keeps distracting during
> debugging test failures.

[ovirt-devel] Re: [VDSM] Travis builds still fail on .coverage rename

2018-07-06 Thread Edward Haas
I do not know if it is relevant or not, but the tests that travis runs for
master are taken from the 4.2 branch.
OVS tests are now running using pytest.


On Fri, Jul 6, 2018 at 12:51 AM, Nir Soffer  wrote:

>
>
> On Thu, Jul 5, 2018 at 10:55 PM Nir Soffer  wrote:
>
>> On Thu, Jul 5, 2018 at 5:53 PM Nir Soffer  wrote:
>>
>>> On Thu, Jul 5, 2018 at 5:43 PM Dan Kenigsberg  wrote:
>>>
 On Thu, Jul 5, 2018 at 2:52 AM, Nir Soffer  wrote:
 > On Wed, Jul 4, 2018 at 1:00 PM Dan Kenigsberg 
 wrote:
 >>
 >> On Wed, Jul 4, 2018 at 12:48 PM, Nir Soffer 
 wrote:
 >> > Dan, travis build still fail when renaming coverage file even after
 >> > your last patch.
 >> >
 >> >
 >> >
 >> > ...SS.SS
 
 
 .SS.
 .S.SS...
 .SS.SS..
 ..S...SSS...S.S.
 S...
 .SSS..S.SS..
 
 
 
 >> > 
 --
 >> > Ran 1267 tests in 99.239s
 >> > OK (SKIP=63)
 >> > [ -n "$NOSE_WITH_COVERAGE" ] && mv .coverage .coverage-nose-py2
 >> > make[1]: *** [check] Error 1
 >> > make[1]: Leaving directory `/vdsm/tests'
 >> > ERROR: InvocationError: '/usr/bin/make -C tests check'
 >> >
 >> > https://travis-ci.org/oVirt/vdsm/jobs/399932012
 >> >
 >> > Do you have any idea what is wrong there?
 >> >
 >> > Why we don't have any error message from the failed command?
 >>
 >> No idea, nothing pops to mind.
 >> We can revert to the sillier [ -f .coverage ] condition instead of
 >> understanding (yeah, this feels dirty)
 >
 >
 > Thanks, your patch (https://gerrit.ovirt.org/#/c/92813/) fixed this
 > failure.
 >
 > Now we have failures for the pywatch_test, and some network
 > tests. Can someone from network look at this?
 > https://travis-ci.org/nirs/vdsm/builds/400204807

 https://travis-ci.org/nirs/vdsm/jobs/400204808 shows

   ConfigNetworkError: (21, 'Executing commands failed:
 ovs-vsctl: cannot create a bridge named vdsmbr_test because a bridge
 named vdsmbr_test already exists')

 which I thought was limited to dirty ovirt-ci jenkins slaves. Any idea
 why it shows here?

>>>
>>> Maybe one failed test leave dirty host to the next test?
>>>
>>
> network tests fail now only on CentOS now.
>
>
>>
>>>
 py-watch seems to be failing due to missing gdb on the travis image
>>>
>>>
 cmdutils.py151 DEBUG./py-watch 0.1 sleep 10 (cwd
 None)
 cmdutils.py159 DEBUGFAILED:  = 'Traceback
 (most recent call last):\n  File "./py-watch", line 60, in \n
   dump_trace(watched_proc)\n  File "./py-watch", line 32, in
 dump_trace\n\'thread apply all py-bt\'])\n  File
 "/usr/lib64/python2.7/site-packages/subprocess32.py", line 575, in
 call\np = Popen(*popenargs, **kwargs)\n  File
 "/usr/lib64/python2.7/site-packages/subprocess32.py", line 822, in
 __init__\nrestore_signals, start_new_session)\n  File
 "/usr/lib64/python2.7/site-packages/subprocess32.py", line 1567, in
 _execute_child\nraise child_exception_type(errno_num,
 err_msg)\nOSError: [Errno 2] No such file or directory: \'gdb\'\n';
  = 1

>>>
>>> Cool, easy fix.
>>>
>>
>> Fixed by https://gerrit.ovirt.org/#/c/92846/
>>
>
> Fedora 28 build is green with this change:
> https://travis-ci.org/nirs/vdsm/jobs/400549561
>
>
>
> ___ summary 
> 
>   tests: commands succeeded
>   storage-py27: commands succeeded
>   storage-py36: commands succeeded
>   lib-py27: commands succeeded
>   lib-py36: commands succeeded
>   network-py27: commands succeeded
>   network-py36: commands succeeded
>   virt-py27: commands succeeded
>   virt-py36: commands succeeded
>   congratulations :)
>
>
>
>>
>>>
>>>
 Nir, could you remind me what is "ERROR: InterpreterNotFound:
 python3.6" and how can we avoid it? it keeps distracting during
 debugging test failures.

>>>
>>> We can avoid it in travis using env matrix.
>>>
>>> Currently we run "make check" which run all the the tox envs
>>> (e.g. storage-py27,storage-py36) regardless of the build type. This is
>>> good
>>> for manual usage when you don't know which python version is

[ovirt-devel] Re: [VDSM] Travis builds still fail on .coverage rename

2018-07-05 Thread Nir Soffer
On Thu, Jul 5, 2018 at 10:55 PM Nir Soffer  wrote:

> On Thu, Jul 5, 2018 at 5:53 PM Nir Soffer  wrote:
>
>> On Thu, Jul 5, 2018 at 5:43 PM Dan Kenigsberg  wrote:
>>
>>> On Thu, Jul 5, 2018 at 2:52 AM, Nir Soffer  wrote:
>>> > On Wed, Jul 4, 2018 at 1:00 PM Dan Kenigsberg 
>>> wrote:
>>> >>
>>> >> On Wed, Jul 4, 2018 at 12:48 PM, Nir Soffer 
>>> wrote:
>>> >> > Dan, travis build still fail when renaming coverage file even after
>>> >> > your last patch.
>>> >> >
>>> >> >
>>> >> >
>>> >> >
>>> ...SS.SS.SS..S.SSSS.SSS...SSS...S.S.SSSS..S.SS..
>>> >> >
>>> --
>>> >> > Ran 1267 tests in 99.239s
>>> >> > OK (SKIP=63)
>>> >> > [ -n "$NOSE_WITH_COVERAGE" ] && mv .coverage .coverage-nose-py2
>>> >> > make[1]: *** [check] Error 1
>>> >> > make[1]: Leaving directory `/vdsm/tests'
>>> >> > ERROR: InvocationError: '/usr/bin/make -C tests check'
>>> >> >
>>> >> > https://travis-ci.org/oVirt/vdsm/jobs/399932012
>>> >> >
>>> >> > Do you have any idea what is wrong there?
>>> >> >
>>> >> > Why we don't have any error message from the failed command?
>>> >>
>>> >> No idea, nothing pops to mind.
>>> >> We can revert to the sillier [ -f .coverage ] condition instead of
>>> >> understanding (yeah, this feels dirty)
>>> >
>>> >
>>> > Thanks, your patch (https://gerrit.ovirt.org/#/c/92813/) fixed this
>>> > failure.
>>> >
>>> > Now we have failures for the pywatch_test, and some network
>>> > tests. Can someone from network look at this?
>>> > https://travis-ci.org/nirs/vdsm/builds/400204807
>>>
>>> https://travis-ci.org/nirs/vdsm/jobs/400204808 shows
>>>
>>>   ConfigNetworkError: (21, 'Executing commands failed:
>>> ovs-vsctl: cannot create a bridge named vdsmbr_test because a bridge
>>> named vdsmbr_test already exists')
>>>
>>> which I thought was limited to dirty ovirt-ci jenkins slaves. Any idea
>>> why it shows here?
>>>
>>
>> Maybe one failed test leave dirty host to the next test?
>>
>
network tests fail now only on CentOS now.


>
>>
>>> py-watch seems to be failing due to missing gdb on the travis image
>>
>>
>>> cmdutils.py151 DEBUG./py-watch 0.1 sleep 10 (cwd
>>> None)
>>> cmdutils.py159 DEBUGFAILED:  = 'Traceback
>>> (most recent call last):\n  File "./py-watch", line 60, in \n
>>>   dump_trace(watched_proc)\n  File "./py-watch", line 32, in
>>> dump_trace\n\'thread apply all py-bt\'])\n  File
>>> "/usr/lib64/python2.7/site-packages/subprocess32.py", line 575, in
>>> call\np = Popen(*popenargs, **kwargs)\n  File
>>> "/usr/lib64/python2.7/site-packages/subprocess32.py", line 822, in
>>> __init__\nrestore_signals, start_new_session)\n  File
>>> "/usr/lib64/python2.7/site-packages/subprocess32.py", line 1567, in
>>> _execute_child\nraise child_exception_type(errno_num,
>>> err_msg)\nOSError: [Errno 2] No such file or directory: \'gdb\'\n';
>>>  = 1
>>>
>>
>> Cool, easy fix.
>>
>
> Fixed by https://gerrit.ovirt.org/#/c/92846/
>

Fedora 28 build is green with this change:
https://travis-ci.org/nirs/vdsm/jobs/400549561



___ summary 
  tests: commands succeeded
  storage-py27: commands succeeded
  storage-py36: commands succeeded
  lib-py27: commands succeeded
  lib-py36: commands succeeded
  network-py27: commands succeeded
  network-py36: commands succeeded
  virt-py27: commands succeeded
  virt-py36: commands succeeded
  congratulations :)



>
>>
>>
>>> Nir, could you remind me what is "ERROR: InterpreterNotFound:
>>> python3.6" and how can we avoid it? it keeps distracting during
>>> debugging test failures.
>>>
>>
>> We can avoid it in travis using env matrix.
>>
>> Currently we run "make check" which run all the the tox envs
>> (e.g. storage-py27,storage-py36) regardless of the build type. This is
>> good
>> for manual usage when you don't know which python version is available
>> on a developer machine. For example if I have python 3.7 installed, maybe
>> I like to test.
>>
>> We can change this so we will test only the *-py27 on centos, and both
>> *-py27 and *-py36 on Fedora.
>>
>> We can do the same in ovirt CI but it will be harder, we don't have a
>> declerative
>> way to configure this.
>>
>
> Fixed all builds using --enable-python3:
> https://gerrit.ovirt.org/#/c/92

[ovirt-devel] Re: [VDSM] Travis builds still fail on .coverage rename

2018-07-05 Thread Nir Soffer
On Thu, Jul 5, 2018 at 5:53 PM Nir Soffer  wrote:

> On Thu, Jul 5, 2018 at 5:43 PM Dan Kenigsberg  wrote:
>
>> On Thu, Jul 5, 2018 at 2:52 AM, Nir Soffer  wrote:
>> > On Wed, Jul 4, 2018 at 1:00 PM Dan Kenigsberg 
>> wrote:
>> >>
>> >> On Wed, Jul 4, 2018 at 12:48 PM, Nir Soffer 
>> wrote:
>> >> > Dan, travis build still fail when renaming coverage file even after
>> >> > your last patch.
>> >> >
>> >> >
>> >> >
>> >> >
>> ...SS.SS.SS..S.SSSS.SSS...SSS...S.S.SSSS..S.SS..
>> >> >
>> --
>> >> > Ran 1267 tests in 99.239s
>> >> > OK (SKIP=63)
>> >> > [ -n "$NOSE_WITH_COVERAGE" ] && mv .coverage .coverage-nose-py2
>> >> > make[1]: *** [check] Error 1
>> >> > make[1]: Leaving directory `/vdsm/tests'
>> >> > ERROR: InvocationError: '/usr/bin/make -C tests check'
>> >> >
>> >> > https://travis-ci.org/oVirt/vdsm/jobs/399932012
>> >> >
>> >> > Do you have any idea what is wrong there?
>> >> >
>> >> > Why we don't have any error message from the failed command?
>> >>
>> >> No idea, nothing pops to mind.
>> >> We can revert to the sillier [ -f .coverage ] condition instead of
>> >> understanding (yeah, this feels dirty)
>> >
>> >
>> > Thanks, your patch (https://gerrit.ovirt.org/#/c/92813/) fixed this
>> > failure.
>> >
>> > Now we have failures for the pywatch_test, and some network
>> > tests. Can someone from network look at this?
>> > https://travis-ci.org/nirs/vdsm/builds/400204807
>>
>> https://travis-ci.org/nirs/vdsm/jobs/400204808 shows
>>
>>   ConfigNetworkError: (21, 'Executing commands failed:
>> ovs-vsctl: cannot create a bridge named vdsmbr_test because a bridge
>> named vdsmbr_test already exists')
>>
>> which I thought was limited to dirty ovirt-ci jenkins slaves. Any idea
>> why it shows here?
>>
>
> Maybe one failed test leave dirty host to the next test?
>
>
>> py-watch seems to be failing due to missing gdb on the travis image
>
>
>> cmdutils.py151 DEBUG./py-watch 0.1 sleep 10 (cwd None)
>> cmdutils.py159 DEBUGFAILED:  = 'Traceback
>> (most recent call last):\n  File "./py-watch", line 60, in \n
>>   dump_trace(watched_proc)\n  File "./py-watch", line 32, in
>> dump_trace\n\'thread apply all py-bt\'])\n  File
>> "/usr/lib64/python2.7/site-packages/subprocess32.py", line 575, in
>> call\np = Popen(*popenargs, **kwargs)\n  File
>> "/usr/lib64/python2.7/site-packages/subprocess32.py", line 822, in
>> __init__\nrestore_signals, start_new_session)\n  File
>> "/usr/lib64/python2.7/site-packages/subprocess32.py", line 1567, in
>> _execute_child\nraise child_exception_type(errno_num,
>> err_msg)\nOSError: [Errno 2] No such file or directory: \'gdb\'\n';
>>  = 1
>>
>
> Cool, easy fix.
>

Fixed by https://gerrit.ovirt.org/#/c/92846/


>
>
>> Nir, could you remind me what is "ERROR: InterpreterNotFound:
>> python3.6" and how can we avoid it? it keeps distracting during
>> debugging test failures.
>>
>
> We can avoid it in travis using env matrix.
>
> Currently we run "make check" which run all the the tox envs
> (e.g. storage-py27,storage-py36) regardless of the build type. This is good
> for manual usage when you don't know which python version is available
> on a developer machine. For example if I have python 3.7 installed, maybe
> I like to test.
>
> We can change this so we will test only the *-py27 on centos, and both
> *-py27 and *-py36 on Fedora.
>
> We can do the same in ovirt CI but it will be harder, we don't have a
> declerative
> way to configure this.
>

Fixed all builds using --enable-python3:
https://gerrit.ovirt.org/#/c/92847/

Nir
___
Devel mailing list -- devel@ovirt.org
To unsubscribe send an email to devel-le...@ovirt.org
Privacy Statement: https://www.ovirt.org/site/privacy-policy/
oVirt Code of Conduct: 
https://www.ovirt.org/community/about/community-guidelines/
List Archives: 
https://lists.ovirt.org/archives/list/devel@ovirt.org/message/ZOMIB3FLERFSFFGSVDBOSHOZNJGP65UQ/


[ovirt-devel] Re: [VDSM] Travis builds still fail on .coverage rename

2018-07-05 Thread Nir Soffer
On Thu, Jul 5, 2018 at 5:43 PM Dan Kenigsberg  wrote:

> On Thu, Jul 5, 2018 at 2:52 AM, Nir Soffer  wrote:
> > On Wed, Jul 4, 2018 at 1:00 PM Dan Kenigsberg  wrote:
> >>
> >> On Wed, Jul 4, 2018 at 12:48 PM, Nir Soffer  wrote:
> >> > Dan, travis build still fail when renaming coverage file even after
> >> > your last patch.
> >> >
> >> >
> >> >
> >> >
> ...SS.SS.SS..S.SSSS.SSS...SSS...S.S.SSSS..S.SS..
> >> > --
> >> > Ran 1267 tests in 99.239s
> >> > OK (SKIP=63)
> >> > [ -n "$NOSE_WITH_COVERAGE" ] && mv .coverage .coverage-nose-py2
> >> > make[1]: *** [check] Error 1
> >> > make[1]: Leaving directory `/vdsm/tests'
> >> > ERROR: InvocationError: '/usr/bin/make -C tests check'
> >> >
> >> > https://travis-ci.org/oVirt/vdsm/jobs/399932012
> >> >
> >> > Do you have any idea what is wrong there?
> >> >
> >> > Why we don't have any error message from the failed command?
> >>
> >> No idea, nothing pops to mind.
> >> We can revert to the sillier [ -f .coverage ] condition instead of
> >> understanding (yeah, this feels dirty)
> >
> >
> > Thanks, your patch (https://gerrit.ovirt.org/#/c/92813/) fixed this
> > failure.
> >
> > Now we have failures for the pywatch_test, and some network
> > tests. Can someone from network look at this?
> > https://travis-ci.org/nirs/vdsm/builds/400204807
>
> https://travis-ci.org/nirs/vdsm/jobs/400204808 shows
>
>   ConfigNetworkError: (21, 'Executing commands failed:
> ovs-vsctl: cannot create a bridge named vdsmbr_test because a bridge
> named vdsmbr_test already exists')
>
> which I thought was limited to dirty ovirt-ci jenkins slaves. Any idea
> why it shows here?
>

Maybe one failed test leave dirty host to the next test?


> py-watch seems to be failing due to missing gdb on the travis image


> cmdutils.py151 DEBUG./py-watch 0.1 sleep 10 (cwd None)
> cmdutils.py159 DEBUGFAILED:  = 'Traceback
> (most recent call last):\n  File "./py-watch", line 60, in \n
>   dump_trace(watched_proc)\n  File "./py-watch", line 32, in
> dump_trace\n\'thread apply all py-bt\'])\n  File
> "/usr/lib64/python2.7/site-packages/subprocess32.py", line 575, in
> call\np = Popen(*popenargs, **kwargs)\n  File
> "/usr/lib64/python2.7/site-packages/subprocess32.py", line 822, in
> __init__\nrestore_signals, start_new_session)\n  File
> "/usr/lib64/python2.7/site-packages/subprocess32.py", line 1567, in
> _execute_child\nraise child_exception_type(errno_num,
> err_msg)\nOSError: [Errno 2] No such file or directory: \'gdb\'\n';
>  = 1
>

Cool, easy fix.


> Nir, could you remind me what is "ERROR: InterpreterNotFound:
> python3.6" and how can we avoid it? it keeps distracting during
> debugging test failures.
>

We can avoid it in travis using env matrix.

Currently we run "make check" which run all the the tox envs
(e.g. storage-py27,storage-py36) regardless of the build type. This is good
for manual usage when you don't know which python version is available
on a developer machine. For example if I have python 3.7 installed, maybe
I like to test.

We can change this so we will test only the *-py27 on centos, and both
*-py27 and *-py36 on Fedora.

We can do the same in ovirt CI but it will be harder, we don't have a
declerative
way to configure this.

Nir
___
Devel mailing list -- devel@ovirt.org
To unsubscribe send an email to devel-le...@ovirt.org
Privacy Statement: https://www.ovirt.org/site/privacy-policy/
oVirt Code of Conduct: 
https://www.ovirt.org/community/about/community-guidelines/
List Archives: 
https://lists.ovirt.org/archives/list/devel@ovirt.org/message/UKVQVCL2M6NE5VHZQMIPNYNSREIJOQUQ/


[ovirt-devel] Re: [VDSM] Travis builds still fail on .coverage rename

2018-07-05 Thread Dan Kenigsberg
On Thu, Jul 5, 2018 at 2:52 AM, Nir Soffer  wrote:
> On Wed, Jul 4, 2018 at 1:00 PM Dan Kenigsberg  wrote:
>>
>> On Wed, Jul 4, 2018 at 12:48 PM, Nir Soffer  wrote:
>> > Dan, travis build still fail when renaming coverage file even after
>> > your last patch.
>> >
>> >
>> >
>> > ...SS.SS.SS..S.SSSS.SSS...SSS...S.S.SSSS..S.SS..
>> > --
>> > Ran 1267 tests in 99.239s
>> > OK (SKIP=63)
>> > [ -n "$NOSE_WITH_COVERAGE" ] && mv .coverage .coverage-nose-py2
>> > make[1]: *** [check] Error 1
>> > make[1]: Leaving directory `/vdsm/tests'
>> > ERROR: InvocationError: '/usr/bin/make -C tests check'
>> >
>> > https://travis-ci.org/oVirt/vdsm/jobs/399932012
>> >
>> > Do you have any idea what is wrong there?
>> >
>> > Why we don't have any error message from the failed command?
>>
>> No idea, nothing pops to mind.
>> We can revert to the sillier [ -f .coverage ] condition instead of
>> understanding (yeah, this feels dirty)
>
>
> Thanks, your patch (https://gerrit.ovirt.org/#/c/92813/) fixed this
> failure.
>
> Now we have failures for the pywatch_test, and some network
> tests. Can someone from network look at this?
> https://travis-ci.org/nirs/vdsm/builds/400204807

https://travis-ci.org/nirs/vdsm/jobs/400204808 shows

  ConfigNetworkError: (21, 'Executing commands failed:
ovs-vsctl: cannot create a bridge named vdsmbr_test because a bridge
named vdsmbr_test already exists')

which I thought was limited to dirty ovirt-ci jenkins slaves. Any idea
why it shows here?

py-watch seems to be failing due to missing gdb on the travis image

cmdutils.py151 DEBUG./py-watch 0.1 sleep 10 (cwd None)
cmdutils.py159 DEBUGFAILED:  = 'Traceback
(most recent call last):\n  File "./py-watch", line 60, in \n
  dump_trace(watched_proc)\n  File "./py-watch", line 32, in
dump_trace\n\'thread apply all py-bt\'])\n  File
"/usr/lib64/python2.7/site-packages/subprocess32.py", line 575, in
call\np = Popen(*popenargs, **kwargs)\n  File
"/usr/lib64/python2.7/site-packages/subprocess32.py", line 822, in
__init__\nrestore_signals, start_new_session)\n  File
"/usr/lib64/python2.7/site-packages/subprocess32.py", line 1567, in
_execute_child\nraise child_exception_type(errno_num,
err_msg)\nOSError: [Errno 2] No such file or directory: \'gdb\'\n';
 = 1


Nir, could you remind me what is "ERROR: InterpreterNotFound:
python3.6" and how can we avoid it? it keeps distracting during
debugging test failures.





>
> Nir
___
Devel mailing list -- devel@ovirt.org
To unsubscribe send an email to devel-le...@ovirt.org
Privacy Statement: https://www.ovirt.org/site/privacy-policy/
oVirt Code of Conduct: 
https://www.ovirt.org/community/about/community-guidelines/
List Archives: 
https://lists.ovirt.org/archives/list/devel@ovirt.org/message/DIWMHL7ITMMUM32E6SPG247JFOXH5WRT/


[ovirt-devel] Re: [VDSM] Travis builds still fail on .coverage rename

2018-07-04 Thread Nir Soffer
On Wed, Jul 4, 2018 at 1:00 PM Dan Kenigsberg  wrote:

> On Wed, Jul 4, 2018 at 12:48 PM, Nir Soffer  wrote:
> > Dan, travis build still fail when renaming coverage file even after
> > your last patch.
> >
> >
> >
> ...SS.SS.SS..S.SSSS.SSS...SSS...S.S.SSSS..S.SS..
> > --
> > Ran 1267 tests in 99.239s
> > OK (SKIP=63)
> > [ -n "$NOSE_WITH_COVERAGE" ] && mv .coverage .coverage-nose-py2
> > make[1]: *** [check] Error 1
> > make[1]: Leaving directory `/vdsm/tests'
> > ERROR: InvocationError: '/usr/bin/make -C tests check'
> >
> > https://travis-ci.org/oVirt/vdsm/jobs/399932012
> >
> > Do you have any idea what is wrong there?
> >
> > Why we don't have any error message from the failed command?
>
> No idea, nothing pops to mind.
> We can revert to the sillier [ -f .coverage ] condition instead of
> understanding (yeah, this feels dirty)
>

Thanks, your patch (https://gerrit.ovirt.org/#/c/92813/) fixed this
failure.

Now we have failures for the pywatch_test, and some network
tests. Can someone from network look at this?
https://travis-ci.org/nirs/vdsm/builds/400204807

Nir
___
Devel mailing list -- devel@ovirt.org
To unsubscribe send an email to devel-le...@ovirt.org
Privacy Statement: https://www.ovirt.org/site/privacy-policy/
oVirt Code of Conduct: 
https://www.ovirt.org/community/about/community-guidelines/
List Archives: 
https://lists.ovirt.org/archives/list/devel@ovirt.org/message/AVK2LE7QERKUUI2LOI3OBMQWT3ARTMWW/


[ovirt-devel] Re: [VDSM] Travis builds still fail on .coverage rename

2018-07-04 Thread Dan Kenigsberg
On Wed, Jul 4, 2018 at 12:48 PM, Nir Soffer  wrote:
> Dan, travis build still fail when renaming coverage file even after
> your last patch.
>
>
> ...SS.SS.SS..S.SSSS.SSS...SSS...S.S.SSSS..S.SS..
> --
> Ran 1267 tests in 99.239s
> OK (SKIP=63)
> [ -n "$NOSE_WITH_COVERAGE" ] && mv .coverage .coverage-nose-py2
> make[1]: *** [check] Error 1
> make[1]: Leaving directory `/vdsm/tests'
> ERROR: InvocationError: '/usr/bin/make -C tests check'
>
> https://travis-ci.org/oVirt/vdsm/jobs/399932012
>
> Do you have any idea what is wrong there?
>
> Why we don't have any error message from the failed command?

No idea, nothing pops to mind.
We can revert to the sillier [ -f .coverage ] condition instead of
understanding (yeah, this feels dirty)

>
> Nir
___
Devel mailing list -- devel@ovirt.org
To unsubscribe send an email to devel-le...@ovirt.org
Privacy Statement: https://www.ovirt.org/site/privacy-policy/
oVirt Code of Conduct: 
https://www.ovirt.org/community/about/community-guidelines/
List Archives: 
https://lists.ovirt.org/archives/list/devel@ovirt.org/message/I63M3RDEGHB2XW633LHHZBSZLMBP6A6V/