Re: how 8.1.7.4 calculates the # of 'library cache' latch children ?

From: Alberto Dell'Era <alberto.dellera_at_bigfoot.com>
Date: 28 Nov 2003 05:13:54 -0800
Message-ID: <f4ed41c5.0311280513.7210e876@posting.google.com>

> >Does anybody know how the number of 'library cache' latches are
> >determined by the 8.1.7.4 kernel ?

> The number of latches is determined by the size of the shared pool.

After investigating deeper, I've found this snippet from Steve Adam's site:

http://www.ixora.com.au/q+a/library.htm

"Question: I'd like to know more about _kgl_latch_count . Can you help me in this?

Answer: It sets the number of child library cache latches. The default is the least prime number greater than or equal to cpu_count. The maximum is 67. It can safely be increased to combat library cache latch contention, as long as you stick to prime numbers. However it is only effective if the activity across the existing child library cache latches is evenly distributed as shown in V$LATCH_CHILDREN."

It seems that cpu_count plays the main (or only) role here. In fact, on different instances, all running 8.1.7.4 on Sun Solaris 8, I see

cpu_count=2  #latches=3
cpu_count=10 #latches=11
cpu_count=14 #latches=17

It seems that the dimension of the shared pool doesn't really play any role; e.g. changing the size from 50M to 400M the #latches stays the same.

bye
Alberto

PS #latches is calculated by
select count(*) from v$latch_children where name='library cache'; Received on Fri Nov 28 2003 - 07:13:54 CST