Re: Anyone have DISABLE_OOB=ON set in sqlnet.ora?

From: Jared Still <jkstill_at_gmail.com>
Date: Wed, 5 Aug 2015 09:14:32 -0700
Message-ID: <CAORjz=Pg=MM3dYWHA4JbK3dskiNZodpKJGFTkWtAgjAWcRcBsw_at_mail.gmail.com>

On Wed, Jul 8, 2015 at 6:14 AM, Rich Jesse <rjoralist3_at_society.servebeer.com
> wrote:

> Has anyone used this? I can't find any negatives or side effects on MOS
> about disabling it, but then why does OOB checking exist? It's like the
> equivalent of an appendix in humans...
>

Kind of old post, but replying anyway.

If OOB is disabled, then Oracle has to start polling the client periodically.
(this is default behavior for Windows thick clients BTW)

The frequency of the polls is specified by setting the BREAK_POLL_SKIP parameter.
This tells Oracle to poll the client every N packets.

As you can imagine, this can have some performance implications. A performance hack for this is to increase BREAK_POLL_SKIP to some value much larger than default.
(default is 4)

Doing so can make it very difficult to break out of a query via CTL-C. ( That is what OOB is for)

Read more here:

Oracle Net Performance Tuning (Doc ID 67983.1)

and here:
SQL*Net Break/reset To Client Waits (Doc ID 1463815.1)

Jared Still
Certifiable Oracle DBA and Part Time Perl Evangelist Principal Consultant at Pythian
Pythian Blog http://www.pythian.com/blog/author/still/ Oracle Blog: http://jkstill.blogspot.com Home Page: http://jaredstill.com

--
http://www.freelists.org/webpage/oracle-l

Received on Wed Aug 05 2015 - 18:14:32 CEST