[prev in list] [next in list] [prev in thread] [next in thread] 

List:       pgsql-performance
Subject:    Re: Unaccounted regression from postgresql 11 in later versions
From:       Bob Jolliffe <bobjolliffe () gmail ! com>
Date:       2023-05-31 10:26:58
Message-ID: CACd=f9e1T13ChgDUgURit+JaJa2Mo+CeEyz3xKnJibPfOmr21g () mail ! gmail ! com
[Download RAW message or body]

Wow Michael you are absolutely right.  Turning jit off results in a query
execution about twice as fast as pg11.  That is a huge relief.  I will read
the jit related docs and see if there is anything smarter I should be doing
other than disabling jit entirely, but it works a treat for this query.

Regards
Bob

On Wed, 31 May 2023 at 11:11, Michael Christofides <michael@pgmustard.com>
wrote:

> Does anyone have a theory of why pg15 should behave so differently to pg11
>> here?  Better still, any suggestions for configuration that might make pg15
>> behave more like pg10.  I am really dreading the prospect of stepping our
>> many live implementations back to pg11 :-(.
>>
>
> One major factor here appears to be JIT compilation, which is off by
> default in pg11, but on by default in pg12+.
>
> You can see at the bottom of your slowest query plan that about 233s of
> the 240s are JIT related.
>
> There is good info in the docs about tuning, or turning off, JIT:
> https://www.postgresql.org/docs/current/jit-decision.html
>

[Attachment #3 (text/html)]

<div dir="ltr">Wow Michael you are absolutely right.   Turning jit off results in a \
query execution about twice as fast as pg11.   That is a huge relief.   I will read \
the jit related docs and see if there is anything smarter I should be doing other \
than disabling jit entirely, but it works a treat for this \
query.<div><br></div><div>Regards</div><div>Bob</div></div><br><div \
class="gmail_quote"><div dir="ltr" class="gmail_attr">On Wed, 31 May 2023 at 11:11, \
Michael Christofides &lt;<a \
href="mailto:michael@pgmustard.com">michael@pgmustard.com</a>&gt; \
wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px \
0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr"><div \
class="gmail_quote"><blockquote class="gmail_quote" style="margin:0px 0px 0px \
0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div \
dir="ltr"><div>Does anyone have a theory of why pg15 should behave so differently to \
pg11 here?   Better still, any suggestions for configuration that might make pg15 \
behave more like pg10.   I am really dreading the prospect of stepping our many live \
implementations back to pg11 :-(.<br></div></div></blockquote><div><br></div><div>One \
major factor here appears to be JIT compilation, which is off by default in pg11, but \
on by default in pg12+. <br><br>You can see at the bottom of your slowest query plan \
that about 233s of the 240s are JIT related. <br><br>There is good info in the docs \
about tuning, or turning off, JIT: <a \
href="https://www.postgresql.org/docs/current/jit-decision.html" \
target="_blank">https://www.postgresql.org/docs/current/jit-decision.html</a>  \
</div></div></div> </blockquote></div>



[prev in list] [next in list] [prev in thread] [next in thread]