HPUX Error: 232: Connection reset by peer

[PW HP-UX Support Team]

We've got a server running HPUX 11.11 and a bunch of Oracle 8i and 9i databases. An error appeared in the TNS listener log that the DBA has tossed over the fence to me, but It's seems too generic to make heads or tails of.

NS Primary Error: TNS-12547: TNS:lost contact
NS Secondary Error: TNS-12560: TNS:protocol adapter error
NT Generic Error: TNS-00517: Lost contact
HPUX Error: 232: Connection reset by peer

The Oracle explanation for the TNS error is:

*Cause: Partner has unexpectedly gone away, usually during process startup.
*Action: Investigate partner application for abnormal termination. On an Interchange, this can happen if the machine is overloaded.

From what I can make of it, the issue appears to be on the client end, but the "machine is overloaded" message is what I suspect is leading them to believe it's the HP system, as we've had performance issues on the HP server in the past.

Anyone else experience this type of error?

[Christian Tremblay]

This error means that the connection was reset/lost by the OTHER side of the connection not on the HP side.

The "machine is overloaded" message applies to the remote server.

[Darrel Louis]

Hi Mike,

Do you see any message in the syslog or oracle alert log file, regarding shortage on resources.

Was it one particular database or all?

Darrel

[PW HP-UX Support Team]

That's what I figured as well, that the connection was lost at the other end.

Yes, it is only on one database, and there are no out of the ordinary messages in any of the HPs logs.

[Chan 007]

Hi,

Can you check the log file,

$ORACLE_HOME/network/log/listener.log,

That will help you

Chan

[PW HP-UX Support Team]

No unusual messages in the listener log during the time of the incident.

[PW HP-UX Support Team]

Old thread I never closed. Sorry.