Re: 2-node RAC faster than 6-node!

From: Yong Huang <yong321_at_yahoo.com>
Date: Thu, 3 Apr 2008 11:13:00 -0700 (PDT)
Message-ID: <641514.54132.qm@web80603.mail.mud.yahoo.com>


Dan's explanation is very interesting. Abdul, can you check the values in gv$system_event for name like '%way%', i.e., gc cr or current or grant block 2-way or 3-way, events, after the database has been up and running for a given period of time. Compare the 2-way with 3-way values. Then bring 4 nodes down to have only 2 and check the numbers. Dan's theory says in case of 6 nodes, there's relatively more 3-way traffic, which could slow down the overall performance. Correct, Dan?

Yong Huang

> From: Dan Norris <dannorris_at_dannorris.com>
>
> ... In any case where a
> block is cached in another instance, there are three parties involved: the
> requestor, the holder, and the global resource master for the requested
> block. In a two-node cluster, two of these three parties are the same
> instance and therefore the messaging between them is very fast. When a third
> node is added, there's a statistical probability that these three parties
> will involve all three instances for a significant portion of the
> occurrencesof this situation. In the case where three instances are
> involved,the communications take longer since all messaging has to cross the
> interconnect.



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
Received on Thu Apr 03 2008 - 13:13:00 CDT

Original text of this message