OEM LSNR down/up phantom event....

[bazza]

Ok,
Windows NT,
Oracle 9.2.0.1.0 only installed
920 database + default listener

OEM console 9.2.0.1.0 + OMS etc on a separate machine.

The OMS currently monitors three other 920 databases/listeners on other nodes, with no problems.

On one however, when lsnr down/up event is registered, the event fires straight away and shows the listener is down. It isn't, I can connect via sqlplus remotely. The lsnrctl button on the console even works and shows the listener is running. Checking the listener.log however shows the agent is not running a status check every minute (event scheduled to run every minute)on the listener. Database up/down event works ok.

Tried cleanstarting the agent, no avail.

Tried moving the event to history once its fired and deleting it. Flag goes to green, but if i subsequently take the listener down deliberately, the event doesn't fire!

Oracle have been looking at it for the past week, any ideas out there people? Somehow the agent isn't looking where it should do to pick up the listener (I thought it just used listener.ora).

Any help greatly appreciated!

[bazza]

No replies eh?

For anyone who's interested, worked this one out.

When the agent is started, in the snmp_ro.ora file, there's a parameter called snmp.oralehomelistener. Instead of putting the path name in, its putting 'TNS_HOME' in. Checked for this in environment variables and the registry, but its not there. Manually edited the snmp_ro file enables the event to work normally. But bouncing the agent of course just rewrites that file. Working with oracle to locate where the agent is picking this up from...

Baz