[MDEV-21001] 10.2.28 - Galera Cluster Joiner Fatality Created: 2019-11-06  Updated: 2019-11-18  Resolved: 2019-11-18

Status: Closed
Project: MariaDB Server
Component/s: Galera, Galera SST, mariabackup
Affects Version/s: 10.2.28
Fix Version/s: 10.2.29

Type: Bug Priority: Major
Reporter: Rob Brown Assignee: Unassigned
Resolution: Duplicate Votes: 0
Labels: None
Environment:

CentOS 7.x


Issue Links:
Blocks
is blocked by MDEV-20987 InnoDB fails to start when fts table ... Closed
Duplicate
is duplicated by MDEV-20987 InnoDB fails to start when fts table ... Closed

 Description   

If wsrep_on=ON then it's impossible for one node to "join" a galera cluster. Both Donor and Joiner immediately hang rather permanently if attempted.



 Comments   
Comment by Rob Brown [ 2019-11-06 ]

Even if I delete everything and start with a fresh naked database without any tables or anything, the second node refuses to join. I was able to use "galera_new_cluster" to spawn a solo node, but it refuses to start if there is any old tables or entries.

Comment by Rob Brown [ 2019-11-06 ]

If I downgrade from 10.2.28 to 10.2.27 then everything suddenly works perfectly again, so this is only a regression issue.

Comment by Rob Brown [ 2019-11-06 ]

Many people are complaining about the same problem I've had:

https://forums.cpanel.net/threads/mariadb-error-after-update-10-2-28.662609/

Comment by Rob Brown [ 2019-11-06 ]

Oh! I believe this issue is related to MDEV-20987 already known.

Comment by Rob Brown [ 2019-11-06 ]

This appears to be fixed in 10.2.29.

Generated at Thu Feb 08 09:03:50 UTC 2024 using Jira 8.20.16#820016-sha1:9d11dbea5f4be3d4cc21f03a88dd11d8c8687422.