Re: Giant TOAST tables due to many almost empty pages - Mailing list pgsql-bugs

From Rumko
Subject Re: Giant TOAST tables due to many almost empty pages
Date
Msg-id hsf4nh$3g6$1@dough.gmane.org
Whole thread Raw
In response to Giant TOAST tables due to many almost empty pages  (Rumko <rumcic@gmail.com>)
Responses Re: Giant TOAST tables due to many almost empty pages  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-bugs
Tom Lane wrote:
<snip>
>
> There's something extremely wacko about that vacuum output.  A toast
> table should have few, if any, rows that short.  And it's impossible
> to believe there's no free space at all in the table, especially since
> 122*3259181 bytes is still quite a lot less than 3259181 pages.  Are
> you running with any nonstandard storage or vacuum parameters?
>
> regards, tom lane
>

Regarding storage paramaters, you mean ALTER TABLE x SET STORAGE...? Then no.

Only autovacuum settings are changed to be a bit more aggressive, vacuum
settings haven't been touched:
# select name,setting from pg_settings where source not in
('default','override') AND name LIKE '%vacuum%';
              name              | setting
--------------------------------+---------
 autovacuum_max_workers         | 4
 autovacuum_naptime             | 45
 autovacuum_vacuum_cost_delay   | 15
 autovacuum_vacuum_cost_limit   | 2000
 autovacuum_vacuum_scale_factor | 0.1
 autovacuum_vacuum_threshold    | 75
(6 rows)

pgsql-bugs by date:

Previous
From: Tom Lane
Date:
Subject: Re: Giant TOAST tables due to many almost empty pages
Next
From: Russell Smith
Date:
Subject: Re: pg_restore ignores -C when using a restore list -L