Thank your William for your help.
After 1 week, I do not see any improvement. Worse, I have to manually run
"ANALYSE" daily.
Here is my config :
My bloat for this table is 1.04151.
What does it mean exactly ?
Why do we have to analyse the table daily ? Is it normal ?
Thank you so much for y
Hello Ben,
Looks like you need to tune autovacuum to be more aggressive. Make sure
autovacuum=ON (the default), increase autovacuum_max_workers (at least 1
per database, more if autovacuum is falling
behind), autovacuum_vacuum_scale_factor to be ~half of the default and can
be set per table to be
ben.play wrote:
> Thank you for your quick answer !
>
> And ... you are a genius :)
Yeah, i know ;-)
>
> A simple "analyse
> " resolved my problem.
> Do We have to do it regularly ?
it's running regulary (via vacuum-process), but you can (and should) run
it after mass data changes.
An
Thank you for your quick answer !
And ... you are a genius :)
A simple "analyse
" resolved my problem.
Do We have to do it regularly ?
Thank you a lot !
--
View this message in context:
http://postgresql.nabble.com/My-index-doesn-t-write-anymore-but-read-tp5849689p5849699.html
Sent from
ben.play wrote:
> Hi all,
>
> We have a large database with postgre 9.3 (>300 Gb) and our queries are
> (really) long for 6 days without changing anything.
>
> After seeing our log, I saw that the same query on an old data was quick but
> the same query with new data are really slow.
please
Hi all,
We have a large database with postgre 9.3 (>300 Gb) and our queries are
(really) long for 6 days without changing anything.
After seeing our log, I saw that the same query on an old data was quick but
the same query with new data are really slow.
Let me show you a sample of my query on
