Re: Oracle Apps dying (Production)

From: Bill Verzal (BVerzal@KOMATSUNA.COM)
Date: Fri Apr 11 2003 - 10:16:17 EDT


Is this a 2 or 3-tier client-server system ? If so, check the APP servers.
Perhaps they are not playing nice.

BV
___________________________________________
Bill Verzal
AIX Admin/Komatsu America
(847) 970-3726 (v)

                      Theresa Sarver
                      <tsarver.IFMC@SDP To: aix-l@Princeton.EDU
                      S.ORG> cc:
                      Sent by: IBM AIX Subject: Oracle Apps dying (Production)
                      Discussion List
                      <aix-l@Princeton.
                      EDU>

                      04/11/2003 09:05
                      AM
                      Please respond to
                      IBM AIX
                      Discussion List

Hi all;

We have a problem and I'm not real sure how to go about resolving it...

Environment:
p630
AIX 51 ML3
Oracle 8174

There was a power outage Tuesday night that crashed the server. Ever since
then the Oracle Apps have been spontaneously dying on our AIX5 box - with
Shared Pool errors. Apparently the Shared Pool (Memory?) is filling up and
code can't be aged in/out. If the DBA's try to flush the Shared Pool it
won't let them.

We bounced the box yesterday and that appeared to fix the problem. Then at
8AM this morning the Apps died again.

I have Perf Toolbox on order, but unfortunately don't have it yet. So my
tools are somewhat limited at this point.

Does anyone know what I can do from an S/A point of view to see if Oracle
is leaving Shared Memory segments stranded out there in UNIX-land? The
output of an ipcs looks correct, showing all 3 databases.

root@c2qnu01:/>ipcs -am
IPC status from /dev/mem as of Fri Apr 11 09:02:40 CDT 2003
T ID KEY MODE OWNER GROUP CREATOR CGROUP
NATTCH
  SEGSZ CPID LPID ATIME DTIME CTIME
Shared Memory:
m 0 0x580008d1 --rw-rw-rw- root system root system
0 13
4217728 8558 65298 14:35:20 14:35:20 12:30:11
m 1 0x0d002965 --rw-rw-rw- root system root system
0
   1440 16258 61798 8:54:30 8:54:40 12:31:08
m 2 0xffffffff --rw-rw---- root system root system
0
   4096 24008 24008 9:01:15 9:01:15 12:31:16
m 3 0xf9ce4d9c --rw-r----- oracle dba oracle dba
0 84
7101952 23496 58914 9:02:28 9:02:33 12:31:17
m 4 0xd6156140 --rw-r----- oracle dba oracle dba
0 11
6891648 22984 58912 9:00:00 9:00:01 12:31:36
m 5 0xbfca3970 --rw-r----- oracle dba oracle dba
0 9
8807808 31012 52856 9:00:01 9:00:01 12:31:43
m 6 0x7000304b --rw------- dbtuner dba dbtuner dba
0
 879776 65298 65546 14:35:20 no-entry 14:35:20

Anyone got any ideas???

Thanks for your time and assistance;
Theresa



This archive was generated by hypermail 2.1.7 : Wed Apr 09 2008 - 22:16:44 EDT