I dove into the script and figure out what it needed.
The latest documentation I can find (Percona-XtraDB-Cluster-5.7.18-29.20.pdf), in 29.3.4 of “Manual Configuration” of ProxySQL, says to add this entry to the scheduler:
INSERT INTO scheduler(id,interval_ms,filename,arg1,arg2,arg3,arg4) VALUES(1,‘10000’,‘/usr/bin/proxysql_galera_checker’,‘127.0.0.1’,‘6032’,‘0’,‘/var/lib/proxysql/proxysql_galera_checker.log’);
That used to be correct, but the script has left the documentation behind, and now requires different arguments, specifically:
Usage: $0 <hostgroup_id write> [hostgroup_id read] [number writers] [writers are readers 0|1] [log_file]
- HOSTGROUP WRITERS (required) (0…) The hostgroup_id that contains nodes that will server ‘writes’
- HOSTGROUP READERS (optional) (0…) The hostgroup_id that contains nodes that will server ‘reads’
- NUMBER WRITERS (optional) (0…) Maximum number of write hostgroup_id node that can be marked ONLINE
When 0 (default), all nodes can be marked ONLINE
- WRITERS ARE READERS (optional) (0|1) When 1 (default), ONLINE nodes in write hostgroup_id will prefer not
to be ONLINE in read hostgroup_id
- LOG_FILE (optional) file logfile where node state checks & changes are written to (verbose)
So, the documentation should now say to add this entry to the scheduler, for the example setup they walk through in “Manual Configuration”:
INSERT INTO scheduler(id,active,interval_ms,filename,arg1,arg2,arg3,arg4,arg5) VALUES(1,‘1’,‘10000’,‘/usr/bin/proxysql_galera_checker’,‘0’,‘-1’,‘0’,‘1’,‘/var/lib/proxysql/proxysql_galera_checker.log’);
The only other thing I had to do was update /etc/proxysql-admin.cnf with the password for ‘admin’, that I’d set in ProxySQL.
It’s working like a charm now. SOLVED.