The VMware VirtualCenter Server service 2.5.0.64192 terminated unexpectedly

[ken123]

VC service was terminated today, we found error from event viewer log

Event Id: 1000

Source: VMware VirtualCenter Server

Description: An unrecoverable problem has occurred, stopping the VMware VirtualCenter service. Check database connectivity before restarting. Error: Error[VdbODBCError] (-1) ODBC error: (08S01) - [ODBC][Ora]ORA-12571: TNS:

There is similar error in VC2.01, but we are using VC2.5, any patches for VC2.5?

[dstan]

We now have the same issue. We have been running for over a year using Oracle 9i RAC and Friday we moved to a 10G RAC and ever since we have had the VC service stop randomly with the same error you provided.

ORA-12571: tns:packet writer failure

We are currently looking into it and will I will post here if we figure t out - but if you already did can you post your solution here for us and others?

Thanks!

[atbnet]

Has the VC service completley stopped since you upgraded or does it just stop intermittently after working for a while allowing connections?

Ive not familiar with oracle but can you debug log it to see all the connections and success/failure reasons for a login attempt for VC service via the ODBC connection?

Andy, VCP,

If you found this information useful please award points using the buttons at the top of the page accordingly.

[dstan]

It stops intermittently after running for a while. I had my DBA's look at the Oracle logs to see if they could see anything, and so far no - the only error we can find anywhere is in the VI logs with the error code posted above.

We did not have this issue on 9i - it is only on 10g that we experience this.

[alex555550]

Hi,

what version is the oracle client? Have you updates it also? Have you checked the settings in the ODBC settings?

[dstan]

I appreciate the suggestions but we have tried all the simple stuff like that we can think of. The real issue is that the 12571 error is such a generic Oracle error that it could be almost anything. The machines are on the same subnet so we aren't firewalled and the network shouldn't be an issue but that is about the last thing we need to check. Simply because the error code given basically states that the DB was not available at the time it tried so it quits.

For the time being we set the service to restart immediately after it goes down so at the very least it should go back up pretty quickly, but it will still interrupt backups or anything that might be going. We did a trace in the Oracle client and are currently going through that to see if we can find anything.

To answer your questions we checked all updates to the client (10.2.0.4) and rebuilt the system DSN to ensure it was using the 10.2.0.4 ODBC client - what specific settings would you suggest I check in the ODBC client that would be different between 9i and 10g? Realize that this ran for a year with no issues on 9i - we just don't want to fail back to it if we don't have to.