Details
-
Bug
-
Status: Closed (View Workflow)
-
Major
-
Resolution: Fixed
-
2.5.10
-
Ubuntu 20.04
Installed from MariaDB 10.5 repo
Set drain state via MaxGUI
-
MXS-SPRINT-137
Description
When using the Galera Monitor, you are able to update a node to the Drain state when it is the Master server.
I have a 3 node cluster with a Read-Write-Split service. When you update the state of the master server to Maintenance, another server is promoted to master. However, when you update the state of the master server to Drain, that server remains as master and no new connections can be established to that cluster.
Requested feature to either:
- Disallow Galera master to be set to Drain state
- When Galera master is set to Drain state, promote new master is available
Related logs (mariadb-02 is the master at this time):
name: [mariadb-01] status: [Slave, Synced, Running] state: [IN_USE] last opened at: [Mon May 3 17:13:28 2021] last closed at: [not closed] last close reason: [] num sescmd: [0]
|
name: [mariadb-02] status: [Draining, Master, Synced, Running] state: [NOT_IN_USE] last opened at: [not opened] last closed at: [not closed] last close reason: [] num sescmd: [0]
|
name: [mariadb-03] status: [Slave, Synced, Running] state: [IN_USE] last opened at: [Mon May 3 17:13:28 2021] last closed at: [not closed] last close reason: [] num sescmd: [0]
|
2021-05-03 17:13:42 error : (17675469) [mariadbclient] Routing the query failed. Session will be closed.
|
2021-05-03 17:13:42 warning: (17675538) [readwritesplit] (Read-Write-Service) [Read-Write-Service] Write query received from user@::ffff:192.168.1.137. Could not find a valid master connection. Closing client connection.
|
2021-05-03 17:13:42 error : (17675538) [readwritesplit] (Read-Write-Service) Could not find valid server for target type TARGET_MASTER (COM_STMT_EXECUTE: ), closing connection.
|
Screenshot of MaxGUI able to set Master server to Drain:
