Re: Doubt about AccessExclusiveLock in ALTER TABLE .. SET ( .. );
От | Michael Paquier |
---|---|
Тема | Re: Doubt about AccessExclusiveLock in ALTER TABLE .. SET ( .. ); |
Дата | |
Msg-id | CAB7nPqQyCT08NQtOrwiEZf5BsgytKrve9EkvMSn8y0_O39C0OQ@mail.gmail.com обсуждение исходный текст |
Ответ на | Re: Doubt about AccessExclusiveLock in ALTER TABLE .. SET ( .. ); (Fabrízio de Royes Mello <fabriziomello@gmail.com>) |
Ответы |
Re: Doubt about AccessExclusiveLock in ALTER TABLE .. SET (
.. );
|
Список | pgsql-hackers |
On Thu, Aug 6, 2015 at 3:06 AM, Fabrízio de Royes Mello wrote: > On Wed, Aug 5, 2015 at 9:31 AM, Robert Haas wrote: >> Agreed. I think we're making a mountain out of a molehill here. As >> long as the locks that are actually used are monotonic, just use > and >> stick a comment in there explaining that it could need adjustment if >> we use other lock levels in the future. I presume all the lock-levels >> used for DDL are, and will always be, self-exclusive, so why all this >> hand-wringing? >> > > New version attached with suggested changes. Thanks! +# SET autovacuum_* options needs a ShareUpdateExclusiveLock +# so we mix reads with it to see what works or waits s/needs/need/ and I think you mean mixing "writes", not "reads". Those are minor things though, and from my point of view a committer can look at it. Regards, -- Michael
В списке pgsql-hackers по дате отправления: