Working with an initially simple sharding with 10 servers/backends, I got the following variables added to a basic configuration file (/etc/maxscale.cnf):
The refresh_interval configured as 5 seconds is too low for the number of database schemas which are going to be added and removed from the shardings. I'm planning to increase that value as I see lots of resources being spent by that activity.
The primary goal here is that I can't see any events being added to the log, even enabling the log-priority info, I can't see any mapping refresh, any re-mapping due to a lousy database context change (when you do a bad USE <non-existing-db>.
Is it possible to add that information over a notice class in the log file?