Robert Haas wrote: > On Thu, Nov 20, 2014 at 3:44 PM, Josh Berkus <[email protected]> wrote: > >> So that we can have two ways to lose data? > >> > >> Forbidding connections to a database doesn't prevent XID or MXID > >> wraparound. > > > > It does prevent the user from doing anything about it, though, since > > they can't manually vacuum template0 without knowing unpublished hackery. > > True. I don't know what to do about that. Do you?
Maybe tweak autovacuum so that it vacuum-freezes the non-connectable template databases when they are multixact_freeze_min_age old -- or something similar. That would cause the multixact age to go down to zero for those databases with enough frequency. -- Álvaro Herrera http://www.2ndQuadrant.com/ PostgreSQL Development, 24x7 Support, Training & Services -- Sent via pgsql-hackers mailing list ([email protected]) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers
