Closed carlcsaposs-canonical closed 4 weeks ago
potential cause: ERROR unit.mysql/0.juju-log Cluster upgrade failed, ensure pre-upgrade checks are ran first.
Tried with pre-upgrade-check before juju refresh
Result: blocked status upgrade failed. Check logs for rollback instruction
pre-upgrade-debug-log.txt pre-upgrade-debug-log-filtered.txt
Encountered the same issue in a deployment with 7 mysql servers. 5 out of the 7 failed to recover after a machine reboot with the same error.
Complete debug log: debug-log.log Each failing mysql server logs: cinder-mysql.log heat-mysql.log keystone-mysql.log nova-mysql.log placement-mysql.log
Encountered the same issue in a deployment with 7 mysql servers. 5 out of the 7 failed to recover after a machine reboot with the same error.
@gboutry there's a fix on PR #324, released in edge channel. We are working to promote it to stable.
@gboutry have you had the chance to validate the fix?
I have not seen this issue for a long time, will re-open if seen again.
Considering fixed.
Steps to reproduce
Expected behavior
mysql app upgrades successfully & goes into active state
Actual behavior
mysql app enters blocked status
failed to recover cluster.
Versions
Operating system: Ubuntu 22.04.3 LTS
Juju CLI: 3.2.3-genericlinux-amd64
Juju agent: 3.2.0
Charm revision: 99 before refresh (current 8.0/stable), 109 after refresh (current 8.0/edge)
microk8s: MicroK8s v1.26.9 revision 6059
Log output
Juju debug log: sunbeam-debug-log.txt sunbeam-debug-log-filtered.txt
Additional context
Attempted to reproduce issue encountered by @javacruft