Thanks for yous suggestions Michael and David. On Fri, May 8, 2020 at 4:11 PM Michael Lewis <mle...@entrata.com> wrote:
> autovacuum_naptime being only 5 seconds seems too frequent. A lock_timeout > might be 1-5 seconds depending on your system. Usually, DDL can fail and > wait a little time rather than lock the table for minutes and have all > reads back up behind the DDL. > > Given you have autovacuum_vacuum_cost_limit set to unlimited (seems very > odd), I'm not sure a manual vacuum freeze command on the tables with high > age would perform differently. Still, issuing a vacuum freeze and then > killing the autovacuum process might be worth trying. >
