From: Tom Lane <tgl@sss.pgh.pa.us>
To: xu jian <jamesxu@outlook.com>
cc: "pgsql-admin@postgresql.org" <pgsql-admin@postgresql.org>
Subject: Re: blocking issue when create index concurrently
In-reply-to: 
 <MWHPR20MB1421AEC7CEC67B159AC188F6A19A0@MWHPR20MB1421.namprd20.prod.outlook.com>
References: 
 <MWHPR20MB1421AEC7CEC67B159AC188F6A19A0@MWHPR20MB1421.namprd20.prod.outlook.com>
Comments: In-reply-to xu jian <jamesxu@outlook.com>
	message dated "Wed, 14 Dec 2016 23:21:13 +0000"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <1877.1481823217.1@sss.pgh.pa.us>
Content-Transfer-Encoding: quoted-printable
Date: Thu, 15 Dec 2016 12:33:37 -0500
Message-ID: <1878.1481823217@sss.pgh.pa.us>
Precedence: bulk
Sender: pgsql-admin-owner@postgresql.org

xu jian <jamesxu@outlook.com> writes:
>           I was trying to create 2 indexes concurrently(on different tabl=
e) in same database. I didn't expect the first create index operation would=
 block the other one since they were running on different table.

> However, it is not true. I found one creating index command was blocking =
the other one.

There are (IIRC) two separate points in CREATE INDEX CONCURRENTLY where it
has to wait for all other transactions to exit.  That would include a
transaction running another CREATE INDEX (CONCURRENTLY), even if it's for
a different table.  I think running two at once would end in a deadlock
failure, since they would likely end up waiting for each other.

Usually, if you're using C.I.C., you're trying to minimize the impact
on the rest of the system, so running two at once seems a bit
counterproductive anyway.

			regards, tom lane


--=20
Sent via pgsql-admin mailing list (pgsql-admin@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-admin