DB crashes after about 2 hours

[Former Member]

Hi All,

I am having a problem with our system on our production system with an Oracle 9.2.08 database path level 5 the latest bios and sp level.

The system is memory usage will increase to about 850,000kb and then the system will fall over with this error.

Is alive poll failed on resource OraclePRD920TNSListenerFslmfaorar2 with error code 128.

The amount of time the system stays up can vary a lot from anything from 20 minutes to 2 and a half hours.. The problem seems to have Only happened since the inclusion of SP 1 on our windows 2003 box.

The problem does not appear to be related to Oracle database and more like a problem with the Oracle Listner

Can you please assist us with this?

[former_member618758]

Hi,

I would assume you have Oracle Fail Save, but don't know much more without seening the system or more log details.

The Is alive poll does "lsnrctl stat" to check the status of the

listener. The timeout usually happens due to a heavy load on the

network.

I would suggest first change timeout by Using the Fail Safe Manager.

Fail Safe Manager -> cluster resources -> listener ->

policies tab -> timeout

Please increase your default size (x2)

You can also change so so the "Is Alive" will not

use "Lsnrctl" This can be changed as follows

Fail Safe Manager -> cluster resources -> listener ->

Parameters -> bottom "Use the Listener Control Utility for

"Is Alive" polling" and disable the check.

If you do this "Is Alive" only checks for the resource check

done by MSCS cluster. You do not have to stop the listener, so the

changes can be made online.

Please also check the attached notes for further analysis if needed:.

#681359 failure of db operations with failsafe 3.3.2

#156363 MSCS: NET8 configuration for Oracle

Thanks and good luck,

George