Details
-
Bug
-
Status: Closed (View Workflow)
-
Major
-
Resolution: Duplicate
-
10.4.21, 10.5.13
-
I have a 3 node cluster running on Rocky Linux 8. These servers are my staging environment running in Vultr in the same datacenter. However, this issue also occurs in my production environment with 3 bare metal servers connected to the same switch. I just prefer not to test/experiment on my production systems.I have a 3 node cluster running on Rocky Linux 8. These servers are my staging environment running in Vultr in the same datacenter. However, this issue also occurs in my production environment with 3 bare metal servers connected to the same switch. I just prefer not to test/experiment on my production systems.
Description
More often than not when using mariabackup, the donor is stuck in Donor/Desynced indefinitely even though the joiner has finished the sst and is Synced. Sometimes this issue will not occur, but it rarely works correctly. If I restart mariadb on the stuck Donor, everything comes back up fine and fully synced. I can easily reproduce this issue in both my staging and production environments. Attached are the following:
Donor Logs
Joiner Logs
mariadb config
I brought this up in the codership group. Someone has come up with a potential fix for this issue already.
https://groups.google.com/g/codership-team/c/qOeuIRBcrDA
Attachments
Issue Links
- relates to
-
-
- Closed
-