Re: BUG #17513: recompressing already-compressed data via VACUUM FULL or CLUSTER does not work

Поиск
Список
Период
Сортировка
От David G. Johnston
Тема Re: BUG #17513: recompressing already-compressed data via VACUUM FULL or CLUSTER does not work
Дата
Msg-id CAKFQuwbbDcA8=sXYU6scDG3gZzaXjaz0pCNJoEpE-G4ZCd3Wsg@mail.gmail.com
обсуждение исходный текст
Ответ на BUG #17513: recompressing already-compressed data via VACUUM FULL or CLUSTER does not work  (PG Bug reporting form <noreply@postgresql.org>)
Ответы Re: BUG #17513: recompressing already-compressed data via VACUUM FULL or CLUSTER does not work  (Michael Paquier <michael@paquier.xyz>)
Список pgsql-bugs
On Wed, Jun 8, 2022 at 12:46 PM PG Bug reporting form <noreply@postgresql.org> wrote:
The following bug has been logged on the website:

Bug reference:      17513
Logged by:          Kiriakos Georgiou
Email address:      kg.postgresql@olympiakos.com
PostgreSQL version: 14.3
Operating system:   Linux
Description:       

Per
https://www.postgresql.org/message-id/E1lNKw9-0008DT-1L%40gemulon.postgresql.org
and https://www.enterprisedb.com/blog/configurable-lz4-toast-compression
if you change a column's compression algorithm and run VACUUM FULL or
CLUSTER, the data will be recompressed with the new compression algorithm.
This is does not work for 14.2 and 14.3 (the only two versions I tested).

The described behavior didn't even make it into 14.0; and I have no clue where such a thing would be documented to guess if the lack was intentional or not.  I can confirm that as of today the behavior in HEAD is to not decompress and recompress during rewrite.  Frankly, this seems like the expected behavior absent any documentation to the contrary.  Whether there is a use case for someone to code up a way to get a recompression to happen is another matter.  Oddly, I kinda expected that "update cmdata set f1 = f1;" might do it but alas it does not (...set f1 = f1 || ''... works; I'm not quite sure how much worse that, immediately followed by vacuum full, would be compared to whatever it would take to do it all during the full vacuum...especially if one is changing every single row on the table).

David J.

В списке pgsql-bugs по дате отправления:

Предыдущее
От: Aleš Zelený
Дата:
Сообщение: Re: BUG #17512: Process running query fails with SIGSEV - nodeMemoize.c:349
Следующее
От: Michael Paquier
Дата:
Сообщение: Re: BUG #17513: recompressing already-compressed data via VACUUM FULL or CLUSTER does not work