Skip to:

  1. Skip to Jira Navigation
  2. Skip to Side Navigation
  3. Skip to Main Content

mysql pods are getting OOM killed in group replication with sysbench

Description

I was running some group replication cluster with 3 pods and 20Gb of disk space, mysql pod resources are defaults from cr.yaml:

After that I run this script: https://github.com/Percona-QA/cloud-qa/blob/main/ps-data.sh
two times, first time it will finish ok, but second run will fail and mysql pod will get OOM killed:

mysql log:

I don't have time now to investigate more from mysql side so reporting that it doesn't get forgotten.

This is the cr:

Environment

None

Attachments

1

Smart Checklist

Activity

Show:

Tomislav Plavcic December 28, 2022 at 9:43 AM

This is still visible in 0.4.0:

Also the message in operator logs which repeats although the node-0 is up and running:

For mysql I was using:

And for router:

Tomislav Plavcic October 3, 2022 at 7:19 AM

Possibly related to this change: https://jira.percona.com/browse/K8SPS-113

Details

Assignee

Reporter

Labels

Needs QA

Yes

Fix versions

Affects versions

Priority

Smart Checklist

Created September 27, 2022 at 5:32 PM
Updated February 13, 2025 at 5:13 PM