very common false failure of ovirt-engine_master_check-patch-fc25-x86_64
Description
Activity
Former user September 13, 2017 at 8:18 AM
Patch was merged (thanks , and for reviewing!).
I'm marking this issue as resolved, but let's have a couple of days of patches to see if FB passes regularly.
Former user September 12, 2017 at 4:50 PM
Patch https://gerrit.ovirt.org/#/c/81672/ should run FB on EL7 only (as commented above, it's pointless to run it twice per patch). Pending CI run to see if it does what I intended, and if so, for 's review.
Eyal Edri September 12, 2017 at 3:06 PM
After talking with and we agreed there is actually no need to run findbugs more than once (i.e both on fedora and el7) , as its JAVA code, and os independent.
So we'll try to use the new feature from CI to use ENV variables in check-patch.sh to make sure we only run it once on el7, thanks for for looking into it!
Greg Sheremeta September 12, 2017 at 2:23 PM
Actually, this isn't our typical GWT build death. That one occurs when building the webadmin project, and this occurred two projects sooner, uicommonweb.
I actually saw this failure on one of my patches the other day, and I wondered what it was. A simple rerun fixed it.
It does appear that this findbugs needs more memory. Can we test that?
Martin Perina September 12, 2017 at 1:50 PM
This seems to me like GWT build issue, Greg could you please take a look?
Eyal Edri September 12, 2017 at 1:36 PM
not sure there is anything we can help as infra, the VMs should have enough memory as they already build engine, so this sounds like a maven configuration needs to be done
inside the automation/* scripts.
We can leave this ticket open until engine maintainers can verify the current maven settings are OK.
danken September 12, 2017 at 9:03 AM
I hope Engine experts can help here more than me.
On Tue, Sep 12, 2017 at 11:37 AM, Daniel Belenky (oVirt JIRA)
Former user September 12, 2017 at 8:37 AM
How much memory do you estimate engine check-patch needs?
Details
Assignee
Former userFormer user(Deactivated)Reporter
dankendanken(Deactivated)Blocked By
waiting for engine maintainers replyTime tracking
1h loggedPriority
Medium
Details
Details
Assignee
Former userReporter
dankenBlocked By
Time tracking
Priority
Medium
Like many many other executions of the said job,
http://jenkins.ovirt.org/job/ovirt-engine_master_check-patch-fc25-x86_64/13784/
fails with
[INFO] UICommonWeb ........................................ FAILURE [02:26 min]
...
[ERROR] Failed to execute goal
org.codehaus.mojo:findbugs-maven-plugin:3.0.3:findbugs (default-cli)
on project uicommonweb: Execution default-cli of goal
org.codehaus.mojo:findbugs-maven-plugin:3.0.3:findbugs failed: Java
returned: 137 -> [Help 1]
which according to https://github.com/travis-ci/travis-ci/issues/4488
may be sue to lack of memory.
Please look into it; having people blindly retrigger failed jobs is
bad education.