MySQL Forums
Forum List  »  InnoDB clusters

Re: Innodb Cluster Issues
Posted by: Frederic Descamps
Date: May 16, 2018 11:45AM

Hi Joe,
Thank you for testing InnoDB Cluster.

First of all, you are doing what is needed with dba.rebootClusterFromCompleteOutage() BUT if you want to rejoin them all you need to boot them all too.

If you stop all your node, they Will NEVER restart and recreate the cluster by themselves (or at least not for the moment as this is very dangerous).

Guess that node 1 and 2 can see each others when they boot but node 3 is partitioned , then you will end up with two different clusters.

What you need to do in your case (you have two options):

1. start all nodes, use the shell on one and use dba.rebootClusterFromCompleteOutage()

2. start one node, and bootstrap the cluster from it:
- start mysqld
- set group_replication_bootstrap_group=on
- start group_replication
- set group_replication_bootstrap_group=off
then restart all nodes and they will join the cluster (as it has bootstrapped now)

I also advise you to upgrade to the latest 5.7.22 and use mysqlshell 8.0.11 even with 5.7


Options: ReplyQuote

Written By
May 16, 2018 08:17AM
Re: Innodb Cluster Issues
May 16, 2018 11:45AM

Sorry, you can't reply to this topic. It has been closed.

Content reproduced on this site is the property of the respective copyright holders. It is not reviewed in advance by Oracle and does not necessarily represent the opinion of Oracle or any other party.