Comment 4 for bug 1797367
Revision history for this message
| bugproxy (bugproxy) wrote Comment bridged from LTC Bugzilla | #4 |
© 2004
Canonical Ltd.
•
Terms of use
•
Data privacy
•
Contact Launchpad Support
•
Blog
•
Careers
•
System status
•
35a32c5
(Get the code!)
------- Comment From <email address hidden> 2018-10-12 05:58 EDT-------
Thanks for the update. Please find the response below.
1. The issue is reproducible on below versions of kernel:
4.15.0-36.39
4.15.0-35.38
4.15.0-33.36
4.15.0-20.21
Also the panic is reproducible on z14 machine.
We will re run with kernel 4.15.0-38.41 and update the results here
2. With respect to what userspace application triggering happens during the panic, this issue occurs during stress tests on network bonding and at a high level , the set of operations are as below:
# Create bond for which slaves are already configured
# Try to create bond for a slave with wrong port number. ( Failure test case scenario )
# Delete the created bond
The test infrastructure uses REST requests and we have collected the coredump files through nmcli commands along with other operations. As mentioned already in this bugzilla, since crash dump was huge to be added as bugzilla attachment, results of few crash commands like bt, log only were added in the Attachment.
One other observation is, at the time of the crash observed that the gzip tool is running with high cpu consumption reported through top.
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
191999 root 20 0 2424 1424 1152 R 100.0 0.0 0:37.86 gzip
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
68704 root 20 0 2424 1424 1152 R 93.8 0.0 0:41.97 gzip
Will respond back with testing results on 4.15.0-38.41 and a reduced set of exact test scenario. Meanwhile, please let me know where to upload the dump files and sosreport to help in further analysis