Bug #106430 Timeout on wait for view after joining group

• Submitted: 10 Feb 2022 8:41
• Modified: 17 Feb 2022 14:31
• Reporter: yashar ne
• Status: Can't repeat
• Category: MySQL Server: Group Replication
• Severity: S3 (Non-critical)
• OS: Any (OKD)
• Assigned to: MySQL Verification Team
• CPU Architecture: x86
• Tags: group replication

[10 Feb 2022 8:41] yashar ne

Description:
I have three MySQL instances (8.0.28) A (pod), B (pod), C (VM) (A and B are in different clusters). A bootstraps the group then B and C join the group afterward, all servers are ONLINE. B and C are doing fine. If I delete the pod or stop the MySQL unit, they can re-join the cluster without any problems.

But after a while, A becomes unavailable, I see it is attempting to re-join the group but cannot do so. So I delete the pod and run START GROUP_REPLICATION USER='rpl_user', PASSWORD='blablabla'; manually, I see it can find the other members and even tries to connect to them, but it disconnects all of a sudden with Timeout on wait for view after joining group error.

How to repeat:
I'm not sure how to repeat this, this is probably a network issue but logs are not giving enough information on what was the problem exactly or which variable can I set to increase the timeout

[10 Feb 2022 8:43] yashar ne

logs

Attachment: gr_logs.txt (text/plain), 11.01 KiB.

[17 Feb 2022 14:31] MySQL Verification Team

Hi,

I can not reproduce this. Not on bare metal nor on VM (with all 3 nodes on same VM host and with each node on separate VM host). There are parameters to configure timeout values for the GR that you might want to tune if you have network issues (often reported on azure for e.g.)

kind regards