autovacuum_freeze_max_age

For DBAs and operators

Checked against age(pg_class.relfrozenxid). The forced run shows up as autovacuum: VACUUM public.t (to prevent wraparound) in pg_stat_activity and logs automatic aggressive vacuum to prevent wraparound of table "app.public.t": index scans: 1. It does not cancel itself for conflicting locks, which is why a harmless ALTER TABLE can suddenly queue behind it.

Sizing

• Raising it (e.g. 400M–1B) on high-XID-rate systems reduces how often big tables are rescanned, at the cost of a larger pg_xact and less safety margin.
• Per table you can only set it lower: ALTER TABLE t SET (autovacuum_freeze_max_age = 100000000).
• It caps others: vacuum_freeze_table_age ≤ 95%, vacuum_freeze_min_age ≤ 50%, vacuum_failsafe_age ≥ 105% of it.

Watch

SELECT datname, age(datfrozenxid) FROM pg_database ORDER BY 2 DESC;
SELECT c.oid::regclass, age(c.relfrozenxid), pg_size_pretty(pg_table_size(c.oid))
  FROM pg_class c WHERE c.relkind IN ('r','m','t')
 ORDER BY 2 DESC LIMIT 20;

If the age keeps rising past this value although vacuums run, something holds the xmin horizon back: a long transaction, an abandoned prepared transaction (pg_prepared_xacts), a stale replication slot, or a standby with hot_standby_feedback.

How to change it

Change in postgresql.conf or ALTER SYSTEM, then restart the server.