I was having problems starting up the management agent (10.2.0.5) in a RedHat 5 Linux, I was constantly getting the following error:
2009-12-15 19:19:47,406 Thread-1317072 WARN diagnostics.cpu: nmegam_ActivityManager_init: per-thread cpu usage not available for this platform 2009-12-15 19:19:50,251 Thread-1317072 ERROR pingManager: OMS does not understand the timezone region of the agent. Either start the OMS using the extended list of time zones supported by the database or pick a value of time zone from /u01/app/oracle/product/agent10g/sysman/admin/nsupportedtzs.lst , update the property 'agentTZRegion' in the file /u01/app/oracle/product/agent10g/sysman/config/emd.properties and restart the agent.A value which is around an offset of +00:00 from GMT should be picked. 2009-12-15 19:19:50,251 Thread-1317072 ERROR : Signalled to Exit with status 55. OMS requested shutdown 2009-12-15 19:19:50,251 Thread-1317072 ERROR pingManager: Did not receive valid response to ping "UNK_TZ"
The error suggests to modify the value of agentTZRegion in the emd.properties with any of the values in the nsupportedtzs.lst
I did that, and then I followed the official procedure (Setting the Management Agent Time Zone)
But I was still getting the error until I found the answer in a forum
I just had to remove some files
rm -r $AGENT_HOME/sysman/emd/state/*
rm -r $AGENT_HOME/sysman/emd/collection/*
rm -r $AGENT_HOME/sysman/emd/upload/*
And then the problem was gone! The monitoring agent started up nice and quickly!