https://logs.opendaylight.org/releng/vex-yul-odl-jenkins-1/docs-tox-verify-oxygen/547/tox/tox-docs.log.gz

You can see a bunch of module imports are failing. Likely VM is missing some packages.

klou

For change `https://git.opendaylight.org/gerrit/#/c/66496/` does not have any VM related issue. 

Please refer the errors below:

06:20:37 /w/workspace/docs-tox-verify-nitrogen/docs/release-notes/projects/bgp-ls-pcep.rst:180: ERROR: Unknown target name: "rfc7674 <http://tools.ietf.org/html/rfc7674 >".
06:20:37 /w/workspace/docs-tox-verify-nitrogen/docs/release-notes/projects/bier.rst:63: WARNING: Bullet list ends without a blank line; unexpected unindent.
06:20:37 /w/workspace/docs-tox-verify-nitrogen/docs/release-notes/projects/odlparent.rst:76: ERROR: Unknown target name: "7813: karaf: do not package spring<https://bugs.opendaylight.org/show_bug.cgi?id=7813>".
06:20:37 /w/workspace/docs-tox-verify-nitrogen/docs/developer-guide/alto-developer-guide.rst:162: WARNING: Cannot analyze code. No Pygments lexer found for "yang".
06:20:37 /w/workspace/docs-tox-verify-nitrogen/docs/developer-guide/authentication-and-authorization-services.rst:385: ERROR: Unknown directive type "notes".

Some changes are already merged with incorrect directives. `.. notes:` is incorrect, and should be `.. note:`

hope this helps!  

This may not be an issue with the image since the same image is works well for nitrogen, on the oxygen jobs the issue is seen. 

https://jenkins.opendaylight.org/releng/job/docs-tox-verify-nitrogen/34/console

klou, zxiiro Are there any dependencies which are different for oxygen (or new) and nitrogen?  

I am unaware of any difference in dependencies.

Here is another perfectly valid rst file that is failing coala check: https://git.opendaylight.org/gerrit/#/c/66564/

Maybe you can find something in the coala log to point us to the source of the problem?  https://logs.opendaylight.org/releng/vex-yul-odl-jenkins-1/docs-tox-verify-oxygen/552/tox/tox-coala.log.gz

klou Have to tried to fix the coala errors related to formatting as seen in the console logs and checking if it fixes the issues in question ? IMU these errors are failing the job and not missing dependencies or infra issues for [1.]

[1.]https://logs.opendaylight.org/releng/vex-yul-odl-jenkins-1/docs-tox-verify-oxygen/552/console.log.gz

We should not be finding coala errors from other files other than the actual file(s) submitted in the patch.  We should not be running coala checks on these other files - it does not make sense.  What has changed?

Please note that this problem started happening on Dec 15.  Prior to that, we do not have this problem.

> We should not be finding coala errors from other files other than the actual file(s) submitted in the patch.  We should not be running coala checks on these other files - it does not make sense.  What has changed?

klou Not sure if you understand what the tox verify jobs does. All (linting / shell-checks/ jjb / packer)  verify jobs is checked for the entire project and not just the delta of the change in question. If we run the verify locally (or just on your delta) not taking into considering global configuration etc this could cause potential issues later on. Therefore we have to run the verify on the complete repo not just the delta.

The only issue that is seen on this job now is the timeout, which should be fixed in the change [x.]

"Build timed out (after 10 minutes). Marking the build as failed.". Also there has been no changes to the docs job since we migrated to vexxhost. 

[x.] https://git.opendaylight.org/gerrit/66589

I am pretty sure these coala warning/errors are not a result of the migration to the vexxhost and would recommend fixing the coala errors from the logs (also please make sure the warning are fixed) going forward which should get the job to pass. Also we may want to keep an eye out on how the changes were merged with these errors/warnings in place. 

We could not have merged in so many apparently bad rst files all on one day.  It would appear to me that there is still something wrong in the infrastructure.

After increasing the build timeouts the issue is resolved.

Some of the jobs were seeing build timeout because of slow connection to Nexus and git repositories. 

abelur Is this resolved now or is there still something to do here?

I think this is resolved now.