rollback vs. commit for closing read-only transaction

David Parker wrote:

If an application transaction is known to be read-only, is there any
reason to prefer COMMIT or ROLLBACK for closing that transaction? Would
there be any performance difference between the two commands?

Doesn't matter.

-- 
  Bruce Momjian                        |  http://candle.pha.pa.us
  pgman@candle.pha.pa.us               |  (610) 359-1001
  +  If your life is a hard drive,     |  13 Roberts Road
  +  Christ can be your backup.        |  Newtown Square, Pennsylvania 19073

Probably, turning fsync off would be helpful, since you know it is
read-only.

-----Original Message-----
From: pgsql-general-owner@postgresql.org [mailto:pgsql-general-
owner@postgresql.org] On Behalf Of Bruce Momjian
Sent: Monday, April 25, 2005 3:01 PM
To: David Parker
Cc: postgres general
Subject: Re: [GENERAL] rollback vs. commit for closing read-only
transaction

David Parker wrote:

If an application transaction is known to be read-only, is there any
reason to prefer COMMIT or ROLLBACK for closing that transaction?

Would

there be any performance difference between the two commands?

Doesn't matter.

--
Bruce Momjian                        |  http://candle.pha.pa.us
pgman@candle.pha.pa.us               |  (610) 359-1001
+  If your life is a hard drive,     |  13 Roberts Road
+  Christ can be your backup.        |  Newtown Square, Pennsylvania
19073

---------------------------(end of

broadcast)---------------------------

TIP 1: subscribe and unsubscribe commands go to

majordomo@postgresql.org

"Dann Corbit" <DCorbit@connx.com> writes:

Probably, turning fsync off would be helpful, since you know it is
read-only.

Wouldn't make any difference: a transaction that hasn't modified the
database doesn't bother to write any commit/abort WAL record at all.

regards, tom lane

For portability's sake commit successful transactions and rollback those
that fail.

Rick

pgsql-general-owner@postgresql.org wrote on 04/25/2005 05:53:11 PM: