Enterprise Manager OMS high cpu数据收集

Q1. Was there any resource intensive activities were scheduled at that time ?
Q2. What is the frequency of issue i.e. CPU /mem spike ? Was this spike rauns for very long hours (more than 10 min or so) ?

Q3. What the patch details atside ?
$$ORACLE_HOME/OPatch/opatch lsinventory -detail

Q4. From the given screens/report ,,not sure which PID (process ID) is consuming more resource ? if your admin captured , details about the PIDs etc , please share with us .

Please share the following logs for the timestamp when issue was seen:

a) gc_inst/em/EMGC_OMS1/sysman/log/emoms.log, emoms_pbs.trc, emoms_pbs.log ,.emctl.msg file
b) gc_inst/userprojects/Domain/GCDomain/servers/EMGC_OMS1/logs/EMGC_OMS1.out.*
c) Have you captured any system dump or java thread dump during the CPU or memory spike ? if please share the same
d) Get me SYstem log which has captured the TOP resource (CPU & memory) Consumer PID details

Best Regards
Basu

What to collect when high CPU is seen by a java pid ?
Further if you notice any CPU spike or high CPU BY 12c agent /OMS , gather these details

1) OS command , PID details
$topas

2) If Java PID shown in top CPU consumer list , get thread dump using

$<OMS_HOME>/bin/emctl dump oms
$kill -3 <java pid> > /tmp/threadpid.out — generated threaddump ofpid

 

Comment

*

沪ICP备14014813号

沪公网安备 31010802001379号