Groovy script failed

Description

Hi, I had an ovirt-node-ng build failure and didn't get logs about what
happened due to:

Groovy script failed:

java.lang.NullPointerException: Cannot invoke method child() on null object

Failed job is
http://jenkins.ovirt.org/job/ovirt-node-ng_ovirt-4.1_build-artifacts-el7-x86_64/279/
CAn you please check what happened?

SANDRO BONAZZOLA

ASSOCIATE MANAGER, SOFTWARE ENGINEERING, EMEA ENG VIRTUALIZATION R&D

Red Hat EMEA <https://www.redhat.com/>
<https://red.ht/sig>
TRIED. TESTED. TRUSTED. <https://redhat.com/trusted>

Activity

Show:
Barak Korren
September 19, 2017, 12:51 PM

That failure is just a post-build Groovy script that expects some files to be accessible in the job workspace. The real issues seems to be a few lines before:

Looks like Jenkins is timing out while trying to talk to the slave. But this may be caused by the error seen in the actual VM build process.

Dafna Ron
September 19, 2017, 1:43 PM

All 3 failed tasks ran on the same vm.
I marked the vm as offline: vm0090.workers-phx.ovirt.org
and the reason for offline is the failed job: http://jenkins.ovirt.org/job/ovirt-node-ng_ovirt-4.1_build-artifacts-el7-x86_64/279/

Eyal Edri
October 31, 2017, 2:33 PM

do you still see this error? it might have a problem only on that specific VM

Sandro Bonazzola
November 3, 2017, 10:30 AM

Last failure I seen in that job is on September 18th but this bug is not really about the job failures reason, it is about the data collection after the failure, causing NPE in groovy script.
I would expect Exception is handled correctly and possibly avoided. Looks like a bug in the groovy script.

Barak Korren
November 5, 2017, 9:22 AM

Maybe I did not explain things clearly. The exception you see in the groovy code is the result of the execution slave not being available any more by the time the Groovy code runs. It seem that something that the job was doing caused the slave to lose network connectivity with Jenkins. That is the kind of exception we cannot really anticipate of handle properly.

The groovy exception is not the reason behind longs not being available, the slave not being available to pull logs from is...

In any case the failing run was not kept, so we cannot debug further. Next time please mark it to be kept forever and set the description to clearly link to the relevant Jira ticket.

Closing this ticket with CANNOT-REPRODUCE. Please re-open if the issue re-occurs.

Cannot Reproduce

Assignee

infra

Reporter

Sandro Bonazzola

Blocked By

None

Priority

Medium