Details
-
Bug
-
Status: Closed (View Workflow)
-
Major
-
Resolution: Cannot Reproduce
-
2.2.5
Description
MaxScale 2.2.5, MariaDB 10.4.11. Everything runs in containers.
[X.Y.Z.K]
type=server
address=X.Y.Z.K
port=3306
protocol=MariaDBBackend
[Galera-Monitor]
type=monitor
module=galeramon
servers=${IP_LIST}
user=${MAXSCALE_USER}
passwd=${MAXSCALE_PASSWORD}
monitor_interval=2000
[Read-Write-Service]
type=service
router=readconnroute
servers=${LIST_OF_IPS}
user=${MAXSCALE_USER}
passwd=${MAXSCALE_PASSWORD}
MaxScale seems to handle well the situations where I restart DB nodes. I test it by running DROP TABLE ...; CREATE TABLE ...; INSERT INTO ...; in a loop, and then I run SELECT * FROM ... with another client, also in a loop. The DB node is restarted, it comes back up, it rejoins the cluster. All fine so far.
But for a brief while, I get these errors:
ERROR 1047 (08S01) at line 1: WSREP has not yet prepared node for application use
It would be beneficial if MaxScale could make sure that WSREP is ready for use before marking it as a valid node.