In the pre-k8s MediaWiki appserver world we have periodic jobs (previously crons) that regularly fetch geoip databases from the puppetmasters' volatile directory.

The puppet masters have their own jobs/timers so that the active puppetmaster (ca_server) pulls them centrally from Maxmind and then others that sync the volatile dir between puppet masters.

See T288844 for where we recently added some new additional databases using a new license on production appservers.

As @Joe pointed out we should not close that topic until we talked about how to do the same in the future k8s world.

So here is a new task for that to be linked to it.