autovacuum_freeze_max_age on append-only tables - Mailing list pgsql-general

From senor
Subject autovacuum_freeze_max_age on append-only tables
Date
Msg-id DM8PR01MB7016C7C6DCAA853AACEE9039F7F59@DM8PR01MB7016.prod.exchangelabs.com
Whole thread Raw
Responses Re: autovacuum_freeze_max_age on append-only tables  (Peter Geoghegan <pg@bowt.ie>)
Re: autovacuum_freeze_max_age on append-only tables  (Laurenz Albe <laurenz.albe@cybertec.at>)
List pgsql-general
Hi All,
I'm attempting to mimic a new feature in version 13 where INSERTS will trigger vacuum for an append-only table.

I'm using v11 and configuring autovacuum_freeze_max_age to a value representing some number of minutes worth of inserts
ona table containing the current day events. I'm looking to understand the details of how the vacuum operates and what
toexpect and plan for. I first ran into an issue when a script attempted to alter the table to change the value of
autovacuum_freeze_max_agewhile a vacuum was running. I know there is a lock conflict while the vacuum is running but I
wasunder the impression that autovacuum_vacuum_cost_limit would limit the time blocked. The ALTER hung much longer than
Iexpected.  

I'm apparently needing an education on how this "to avoid wraparound" vacuum differs from any other. I've seen it
referencedas "more aggressive" but I'd like details. An upgrade to 13 is "right around the corner". 

Pointers to documentation I might have missed is be appreciated.

-Senor


pgsql-general by date:

Previous
From: Siddhartha Gurijala
Date:
Subject: Large Data insert on Master server fills up /pgsql base dir when Logical replication (pglogical) is active
Next
From: Peter Geoghegan
Date:
Subject: Re: autovacuum_freeze_max_age on append-only tables