Re: Different compression methods for FPI - Mailing list pgsql-hackers

From Andrey Borodin
Subject Re: Different compression methods for FPI
Date
Msg-id 74CD8839-05DD-40F5-9609-58B79AD2151E@yandex-team.ru
Whole thread Raw
In response to Re: Different compression methods for FPI  (Michael Paquier <michael@paquier.xyz>)
Responses Re: Different compression methods for FPI
Re: Different compression methods for FPI
List pgsql-hackers

> 16 июня 2021 г., в 12:18, Michael Paquier <michael@paquier.xyz> написал(а):
>  Among the
> remaining two I would be tempted to choose LZ4.  That's consistent
> with what toast can use now.

I agree that allowing just lz4 - is already a huge step ahead.
But I'd suggest supporting zstd as well. Currently we only compress 8Kb chunks and zstd had no runaway to fully unwrap
it'spotential. 
In WAL-G we observed ~3x improvement in network utilisation when switched from lz4 to zstd in WAL archive compression.

BTW we could get rid of whole hole-in-a-page thing if we would set lz4 as default. This could simplify FPI code.

Thanks!

Best regards, Andrey Borodin.


pgsql-hackers by date:

Previous
From: Kyotaro Horiguchi
Date:
Subject: Re: Duplicate history file?
Next
From: Kyotaro Horiguchi
Date:
Subject: Re: detailed error message of pg_waldump