This indicates some kind of issue when building the RPMs.
To debug what happened, one should go under logs/delorean_repos.tar.xz then dig down into that tarball until they find the rpmbuild.log.
In this case, we can see that the package failed to build because the unit tests are run during package building, and one failed (yet they passed on the review itself... a mystery for another day). So the package can't be built and all related CI jobs fail.
I realise this isn't related to updating dependencies per se but even if this piece of debugging doesn't make it into the docs, the logs to use to figure out if there's a dependency issue are the same, it's simply the error that will be different. build.log seems to have similar information, in a more compact way - maybe better to check there first and dig into rpmbuild.log if the error wasn't obvious in there?
I just saw a somewhat related problem happen on a patch so, adding a few notes here for reference!
See https:/ /review. openstack. org/#/c/ 409791/ 8 . Most test runs are failing, with something like that in the console log:
2016-12-20 22:39:32.467621 | 2016-12-20 22:39:32,466 ERROR:dlrn- build:cmd failed. See logs at: /opt/stack/ new/delorean/ data/repos/ 1c/07/1c0799616 4fe3861f73a9b4c 483112b28a02be8 f_7082fe70/ tripleoclient:
2016-12-20 22:39:32.468800 | 2016-12-20 22:39:32,467 INFO:dlrn:Received exception Error in build_rpm_wrapper for python-
( http:// logs.openstack. org/91/ 409791/ 8/check/ gate-tripleo- ci-centos- 7-nonha- multinode/ cbdbebc/ console. html#_2016- 12-20_22_ 39_32_467621 )
This indicates some kind of issue when building the RPMs.
To debug what happened, one should go under logs/delorean_ repos.tar. xz then dig down into that tarball until they find the rpmbuild.log.
In this case, we can see that the package failed to build because the unit tests are run during package building, and one failed (yet they passed on the review itself... a mystery for another day). So the package can't be built and all related CI jobs fail.
I realise this isn't related to updating dependencies per se but even if this piece of debugging doesn't make it into the docs, the logs to use to figure out if there's a dependency issue are the same, it's simply the error that will be different. build.log seems to have similar information, in a more compact way - maybe better to check there first and dig into rpmbuild.log if the error wasn't obvious in there?