100% CPU utilized (found due to (LOCAL=NO) )

Question

Hi Friends,
OS: SunOS 5.10
I got an alert like CUP is utilizied 100% and oracle consumed 82%
At 7:00 AM
NPROC USERNAME SWAP RSS MEMORY TIME CPU
190 oracle 25G 24G 38% 234:35:05 82%
When I check at 7:30AM
NPROC USERNAME SWAP RSS MEMORY TIME CPU
161 oracle 22G 22G 35% 230:40:43 13%
How to check what happened at 7:00AM.
OS logs gives some idea?
Suggestions appreciated.
Thanks
KSG
Edited by: KSG on Aug 17, 2012 11:33 AM

Answer 1

The following are the process followed ..
To find the CPU consumption
prstat -s cpu -a -n 8
To find the list of process running on the server
ps -eo pcpu,pid,user,args | sort -k 1 -r | head -40
To get the SID from PID
select sid,serial#
from v$session
where paddr in (select addr from v$process
where background is null
and spid=&Server_process_id)
To get query from SID
select sql.sql_text
from v$session ses, v$sqltext sql
where sql.address=ses.sql_address
and sql.hash_value=ses.sql_hash_value
and sid=&session_id
order by piece
To find the connectivity details
select SID,Serial#,status ,program,module,terminal,to_char(LOGON_TIME,'dd-mm-yyyy hh24:mi:ss') from v$session where sid=62;
Check the connectivity accordingly and analyze.
Now to experts,
How to restrict the number of incoming connectivity so that we can always stay under the CPU threshold?
Any idea?
Regards
KSG
Edited by: KSG on Aug 17, 2012 11:30 AM