Re: BUG #11350: ALTER SYSTEM is not DDL?
От | Amit Kapila |
---|---|
Тема | Re: BUG #11350: ALTER SYSTEM is not DDL? |
Дата | |
Msg-id | CAA4eK1KvYoU+P9m1Y=hWDr-Ob4T+T_yMsf9Shv3Qn+XGGh0vPg@mail.gmail.com обсуждение исходный текст |
Ответ на | Re: BUG #11350: ALTER SYSTEM is not DDL? (Andres Freund <andres@2ndquadrant.com>) |
Список | pgsql-bugs |
On Fri, Sep 19, 2014 at 1:05 PM, Andres Freund <andres@2ndquadrant.com> wrote: > > On 2014-09-19 12:50:24 +0530, Amit Kapila wrote: > > On Fri, Sep 19, 2014 at 9:49 AM, Craig Ringer <craig@2ndquadrant.com> wrote: > > > > > > On 09/05/2014 08:32 PM, Amit Kapila wrote: > > > > > > > > and Alter System affects at system level, both of these can't > > > > be considered as DDL, may be a separate category. > > > > > > Like CREATE USER / CREATE ROLE / CREATE DATABASE ? > > > > > > All those are logged as DDL. > > > > They are different from ALTER SYSTEM in terms that they create/modify > > the object (here object can be any database or cluster object) property. > > > > Can you tell me that if we want to make ALTER SYSTEM as DDL, then > > why SET or CHECKPOINT commands can't be DDL? > > Really? Because neither has persistent effects? I am not sure if persistent effects should be considered while defining type of statement. Statements that define database structure or schema in some way should be considered as DDL statements. I am not able to convince myself that ALTER SYSTEM should be a DDL command, however if you and Stephen feels strongly about it then lets do it that way. I think there is already a patch in this thread which defines it as DDL (though I haven't checked it yet). Do you expect anything more? With Regards, Amit Kapila. EnterpriseDB: http://www.enterprisedb.com
В списке pgsql-bugs по дате отправления: