Re: Best Practice for running vacuums during off hours WAS Re: Autovacuum launcher doesn't notice death of postmaster immediately - Mailing list pgsql-hackers

From Alvaro Herrera
Subject Re: Best Practice for running vacuums during off hours WAS Re: Autovacuum launcher doesn't notice death of postmaster immediately
Date
Msg-id 20070607223531.GC13687@alvh.no-ip.org
Whole thread Raw
In response to Best Practice for running vacuums during off hours WAS Re: Autovacuum launcher doesn't notice death of postmaster immediately  ("Andrew Hammond" <andrew.george.hammond@gmail.com>)
List pgsql-hackers
Andrew Hammond escribió:

> That's a good question. I can't see any reason for a naptime longer
> than 60 seconds either.
> 
> I think very large naptime settings are a symptom of another issue:
> what's the Right Way to defer vacuums until "off hours"? Is that even
> a desirable thing anymore? I don't think it is in the majority of
> cases.
> 
> I originally thought that this was more of a Best Practices issue (ie,
> fix in the docs, not the code), but now I'm wondering if there's much
> call for supporting the idea of being more aggressive with vacuums at
> different times of the day / week / month. Anyone?

That's why the intention is to have the autovacuum scheduling feature
be a mechanism for changing the autovac parameters according to date and
time.  We even have a Google Summer of Code project about that.

-- 
Alvaro Herrera                 http://www.amazon.com/gp/registry/DXLWNGRJD34J
"La virtud es el justo medio entre dos defectos" (Aristóteles)


pgsql-hackers by date:

Previous
From: "Andrew Hammond"
Date:
Subject: Best Practice for running vacuums during off hours WAS Re: Autovacuum launcher doesn't notice death of postmaster immediately
Next
From: Gregory Stark
Date:
Subject: Re: [COMMITTERS] pgsql: Avoid losing track of data for shared tables in pgstats.