0

I built a mariadb galera cluster on CentOS7. Below is the galera node information:

10.200.67.27    MariaDB-Node1
10.200.67.29    MariaDB-Node2
10.200.67.26    MariaDB-Node3

However, MariaDB-Node2 was unexpected stopped. And I found a table named leave_records, the data in the table is not synced. I mean the table records is not exactly the same, even though MariaDB-Node1 and MariaDB-Node3 are still running.

How can I compare the diffrence between them, and resync table data? I need every node have same data records.

Any help is important to me.

Winbert
  • 13
  • 5
  • It should be synched automatically during startup. Notice, that Galera originally was developed to sync InnoDB tables (development versions also have experimental support for MyISAM and ARIA). Check log files and configuration, especially regarding SST, and [attach here](https://serverfault.com/posts/1101902/edit), if there are problems that you don't understand. – Nikita Kipriyanov May 27 '22 at 07:33
  • @Nikita Kipriyanov I found the stopped node cannot start up – Winbert May 27 '22 at 08:41
  • Then how you determined that "table records is not exactly the same", without starting the node? Also, where is the configuration and logs from failed node? – Nikita Kipriyanov May 27 '22 at 10:09
  • See if Percona's pt-table-sync has a mode that works with Galera (aka Percona Xtradb Cluster). – Rick James May 27 '22 at 17:20
  • "not exactly the same" -- What is the evidence? (`SHOW TABLE STATUS` is _not_ sufficient.) – Rick James May 27 '22 at 17:21

0 Answers0