Re: BUG #15840: Vacuum does not work after database stopped for wraparound protection. Database seems unrepearable.

From: Alvaro Herrera <alvherre(at)2ndquadrant(dot)com>
To: Thierry Husson <thusson(at)informiciel(dot)com>
Cc: Andres Freund <andres(at)anarazel(dot)de>, pgsql-bugs(at)lists(dot)postgresql(dot)org
Subject: Re: BUG #15840: Vacuum does not work after database stopped for wraparound protection. Database seems unrepearable.
Date: 2019-06-08 00:25:42
Message-ID: 20190608002542.GA4898@alvherre.pgsql
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-bugs pgsql-hackers

On 2019-Jun-07, Thierry Husson wrote:

> Yes autovacuum is enabled. Aggressiveness was effectively a recent problem I
> had and putting its max_worker to 8 wasn't a solution, there were all busy
> 24/7 and I had to do a daily script to help it. The solution was to push
> vacuum_cost_limit to 2000, since then it works like a charm.

Note the I/O cost balancing thing, which seems to bite many people: if
you raise max_workers without changing cost_delay or cost_limit, it
doesn't have much of an effect, because each worker goes slower to
accomodate. Raising the cost limit (or lowering the cost delay) does
have a useful impact. In pg12 we changed the default cost_delay to 2ms
(from 20ms).

--
Álvaro Herrera https://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services

In response to

Browse pgsql-bugs by date

  From Date Subject
Next Message Andres Freund 2019-06-08 00:26:32 Re: Temp table handling after anti-wraparound shutdown (Was: BUG #15840)
Previous Message Michael Paquier 2019-06-07 23:59:37 Re: Temp table handling after anti-wraparound shutdown (Was: BUG #15840)

Browse pgsql-hackers by date

  From Date Subject
Next Message Andres Freund 2019-06-08 00:26:32 Re: Temp table handling after anti-wraparound shutdown (Was: BUG #15840)
Previous Message Michael Paquier 2019-06-08 00:03:17 Re: Custom table AMs need to include heapam.h because of BulkInsertState