Re: sparc 64 solaris 10 112GB RAM, vip hung

From: Tim Gorman <tim_at_evdbt.com>
Date: Sun, 08 Mar 2009 14:15:59 -0600
Message-ID: <49B4277F.2000705_at_evdbt.com>

What does "ulimit -a" on the UNIX account running the CRS say?

Since you're talking about "nodeapps" in the CRS, the problem has nothing whatsoever to do with the RDBMS or ASM, so don't even bother with "init.ora" parms.

Check the easy stuff first -- look at the process limits?

Ujang Jaenudin wrote:
> lists,
>
> env: crs & db 10.2.0.4
>
> I have added new cpu and memory at each node to 8 cpu (dual core = 16
> oracle cpu_count) with memory 112GB RAM, 3 node RAC.
>
> when starting nodeapps got error [28025][1][ora.af-oradb01.vip]:
> clsrcstartorp: Error with malloc
>
> I'm thinking of due to malloc scanning huge memory and time timeout.
>
> during startup instance database (79GB for SGA) it takes 5 minutes to finish.
>
> does anyone has the recipe to enhance oracle performance for this kind
> of memory size?
>
> and what about ISM can helped on this?
>

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

Received on Sun Mar 08 2009 - 15:15:59 CDT