Re: found xmin from before relfrozenxid on pg_catalog.pg_authid
От | Andres Freund |
---|---|
Тема | Re: found xmin from before relfrozenxid on pg_catalog.pg_authid |
Дата | |
Msg-id | D670D841-C88B-4D28-B852-16ABBDBA5F88@anarazel.de обсуждение исходный текст |
Ответ на | Re: found xmin from before relfrozenxid on pg_catalog.pg_authid ("Nasby, Jim" <nasbyj@amazon.com>) |
Ответы |
Re: found xmin from before relfrozenxid on pg_catalog.pg_authid
|
Список | pgsql-hackers |
On May 27, 2018 9:39:49 AM PDT, "Nasby, Jim" <nasbyj@amazon.com> wrote: >On May 26, 2018, at 1:45 PM, Andres Freund <andres@anarazel.de> wrote: >> Does anybody see a way to not have to remove the .init file? > >How about only keeping the critical info for being able to find >relations in the .init files, and then fully populate the cache by >doing a normal lookup? Since there’s multiple entries needed to >bootstrap things I guess there’d need to be a flag in relcache to >indicate that an existing entry wasn’t fully formed. Then the cache wouldn't have any benefits, no? It's been a while, but last time I checked it does make quite a measurableperformance difference in a new backend. Andres -- Sent from my Android device with K-9 Mail. Please excuse my brevity.
В списке pgsql-hackers по дате отправления: