On 3/8/17 04:12, Okano, Naoki wrote:
> Peter Eisentraut wrote:
>> I have a feeling that this was proposed a few times in the ancient past
>> but did not go through because of locking issues. I can't find any
>> emails about it through. Does anyone remember? Have you thought about
>> locking issues?
> Is this e-mail you are finding?
> https://www.postgresql.org/message-id/20140916124537.GH25887%40awork2.anarazel.de
No, that's not the one I had in mind.
> I am considering to add 'OR REPLACE' clause as a first step.
> At least, I think there is no need to change the locking level when replacing a trigger with 'EXECUTE PROCEDURE'
clause.
> In PostgreSQL, we currently have ShareRowExclusiveLock lock on relation on which trigger is created.
ShareRowExclusiveLockis enough to replace a trigger.
> Also, we currently have RowExclusiveLock on pg_trigger. RowExclusiveLock is enough to replace a trigger, too.
I'm not saying it's not correct. I was just wondering.
--
Peter Eisentraut http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
