[regression] MySQL OCF may choose different nodes as possible masters
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Fuel for OpenStack |
Fix Committed
|
High
|
Bogdan Dobrelya | ||
6.1.x |
Won't Fix
|
High
|
MOS Maintenance | ||
7.0.x |
Won't Fix
|
High
|
MOS Maintenance | ||
8.0.x |
Won't Fix
|
High
|
MOS Maintenance | ||
Mitaka |
Fix Released
|
High
|
Bogdan Dobrelya | ||
Newton |
Fix Committed
|
High
|
Bogdan Dobrelya |
Bug Description
When nodes have the same GTID=UUID:SEQNO, different nodes may report different choosen masters.
In result, several node attept to start mysqld --wsrep-new-cluster and fail the cluster (re)assemble.
Steps to reproduce were given in the Galera reliability testing https:/
There is also a way to reproduce it on a Fuel env given that the node-1, node-2, node-3 deployed as controller nodes:
1) https:/
2) PURGE=true ./vagrant_
3) docker exec -it jepsen bash -c "TESTPROC=mysqld lein test :only jepsen.
Changed in fuel: | |
importance: | Undecided → High |
assignee: | nobody → Bogdan Dobrelya (bogdando) |
milestone: | none → 10.0 |
tags: | added: galera |
tags: | added: tech-debt |
description: | updated |
description: | updated |
tags: | added: on-verification |
Fix proposed to branch: master /review. openstack. org/312622
Review: https:/