Re: [PATCH] Support Int64 GUCs
От | Alexander Korotkov |
---|---|
Тема | Re: [PATCH] Support Int64 GUCs |
Дата | |
Msg-id | CAPpHfdusZDd_6ACM5iyk_RNTJVumZsZv1hmrmS+QuCVva1fESA@mail.gmail.com обсуждение исходный текст |
Ответ на | Re: [PATCH] Support Int64 GUCs (wenhui qiu <qiuwenhuifx@gmail.com>) |
Список | pgsql-hackers |
On Thu, Sep 26, 2024 at 12:30 PM wenhui qiu <qiuwenhuifx@gmail.com> wrote: > I think we need int64 GUCs, due to these parameters( autovacuum_freeze_table_age, autovacuum_freeze_max_age,Whena table age is greater than any of these parameters an aggressive vacuum will be performed,When we implementing xid64, is it still necessary to be in the int range? btw, I have a suggestion to record awarning in the log when the table age exceeds the int maximum. These default values we can set a reasonable values ,forexample autovacuum_freeze_max_age=4294967295 or 8589934592. In principle, even with 64-bit transaction ids we could specify *_age GUCs as int32 with bigger units or as float8. That feels a bit awkward for me. This is why I queried more about Tom's opinion in more details: did he propose to wait with int64 GUCs before we have 64-bit transaction ids, or give up about them completely? Links. 1. https://www.postgresql.org/message-id/3649727.1727276882%40sss.pgh.pa.us ------ Regards, Alexander Korotkov Supabase
В списке pgsql-hackers по дате отправления: