Re: Index rebuild online lock issue

From: John D Parker <orclwzrd_at_yahoo.com>
Date: Thu, 3 Apr 2008 15:27:28 -0700 (PDT)
Message-ID: <572271.49144.qm@web32601.mail.mud.yahoo.com>

Thanks for all the suggestions, I was looking for something a bit more defined than just watch it run for XX hours and you'll see the locking occur. I've got the rebuild online running now on my 9.2.0.5 system, but I've still no idea of how to measure/monitor/quantify the impact of the sync process at the end. On other systems, I've inadvertently brought them to their knees with the sync at the end of rebuild online. Just looking to avoid that this time.

john

Yong Huang <yong321_at_yahoo.com> wrote: John, to "measure and monitor", isn't it just a matter of checking v$lock? Also, you can check what locks are taken during the entire process of rebuilding an index by setting event 10704 at level 10 in your session. The trace file should tell you.

Yong Huang

> From: John D Parker
>
> Ok, I know that at the end of alter index rebuild online, a lock is taken
while
> the index that has been built is synchronized with the current state of the
> world as we know it. My question is this. How does one measure/monitor this
lock?
> I know that on a normal "small" index, this lock is almost undetectable.
> However, on a 4TB OLTP system with a 95GB index on a 384GB table is not
small nor
> will the sync to current state be undetectable. How does one measure/monitor
this
> lock?
> Thanks for any ideas or assistance!
> John

---

You rock. That's why Blockbuster's offering you one month of Blockbuster Total Access, No Cost. http://tc.deals.yahoo.com/tc/blockbuster/text5.com

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




       
---------------------------------
You rock. That's why Blockbuster's offering you one month of Blockbuster Total Access, No Cost.
--
http://www.freelists.org/webpage/oracle-l

Received on Thu Apr 03 2008 - 17:27:28 CDT