Nailgun agent sporadic crashes with core dumps

Bug #1754143 reported by Miroslav Anashkin
8
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Fuel for OpenStack
Incomplete
Medium
MOS Maintenance

Bug Description

Nailgun-agent service has generated core dump file in compute node. Alarm "Core Dump Generated" was detected, at that time.

root@compute-2-11011302:/var/log/crash/cores# ls -ltr
total 7144
-rw-r----- 1 root root 7313121 Jan 13 10:45 core.compute-2-11011302.domain.tld.1515807918.ruby.194984.gz

There is no specific steps to reproduce - the next minute nailgun agent continued working as usual.

Attached is the extraction from . the nailgun-agent.log

No corresponding messages appeared in dmesg, kern.log, etc - looks like the crash was fully handled by the Ruby interpreter itself.

This is customer found issue. The impact is that the crash detecting system monitors the core dump appearance and fires alerts.

Revision history for this message
Miroslav Anashkin (manashkin) wrote :
Changed in fuel:
assignee: nobody → MOS Maintenance (mos-maintenance)
importance: Undecided → Medium
importance: Medium → High
Revision history for this message
Alexander Rubtsov (arubtsov) wrote :

sla2 for 9.0-updates

Changed in fuel:
importance: High → Medium
tags: added: sla2
Revision history for this message
Vladimir Jigulin (vjigulin) wrote :

Cannot reproduce, looks like this is a problem with ruby rexml library or ruby itself. Marking as incomplete until this issue is reproduced again. If so, then please provide output of
lstopo --no-caches --of xml

Changed in fuel:
status: New → Incomplete
Revision history for this message
Miroslav Anashkin (manashkin) wrote :

This may be related to the following bug.
https://bugs.launchpad.net/fuel/+bug/1742886

I am still awaiting the results of the requested command from the problematic environment.

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.