Oracle10g Listener getting hung

[Ankit Grover]

Greetings experts!

We have a IA64 box with HP-UX 11.23 and Oracle10g Rel 2. Lately, I have started a strange behaviour from the listener. It just hangs and I have to kill OS process and then re-start it.

The following are the contents of log file around that period of time:

04-FEB-2009 22:44:49 * (CONNECT_DATA=(SID=trial10g)(CID=(PROGRAM=sqlplus)(HOST=hpux1571)(USER=fnsonlw3))) * (ADDRESS=(PROTOCOL=tcp)(HOST=172.30.0.12)(PORT=59909)) * establish * trial10g * 0
04-FEB-2009 22:44:50 * (CONNECT_DATA=(SID=trial10g)(CID=(PROGRAM=sqlplus)(HOST=hpux1571)(USER=fnsonlw3))) * (ADDRESS=(PROTOCOL=tcp)(HOST=172.30.0.12)(PORT=59910)) * establish * trial10g * 0
04-FEB-2009 22:44:50 * (CONNECT_DATA=(SID=trial10g)(CID=(PROGRAM=sqlplus)(HOST=hpux1571)(USER=fnsonlw3))) * (ADDRESS=(PROTOCOL=tcp)(HOST=172.30.0.12)(PORT=59911)) * establish * trial10g * 0
04-FEB-2009 22:44:50 * (CONNECT_DATA=(SID=trial10g)(CID=(PROGRAM=rtsora@hpux1571)(HOST=hpux1571)(USER=fnsonlw4))) * (ADDRESS=(PROTOCOL=tcp)(HOST=172.30.0.12)(PORT=59912)) * establish * trial10g * 0
04-FEB-2009 22:44:51 * (CONNECT_DATA=(SID=trial10g)(CID=(PROGRAM=rtsora@hpux1571)(HOST=hpux1571)(USER=fnsonlw3))) * (ADDRESS=(PROTOCOL=tcp)(HOST=172.30.0.12)(PORT=59913)) * establish * trial10g * 0
05-FEB-2009 09:12:40 * (CONNECT_DATA=(SID=trial10g)(CID=(PROGRAM=sqlplus)(HOST=hpux1571)(USER=fnsonlw3))) * (ADDRESS=(PROTOCOL=tcp)(HOST=172.30.0.12)(PORT=59914)) * establish * trial10g * 12518
TNS-12518: TNS:listener could not hand off client connection
TNS-12547: TNS:lost contact
TNS-12560: TNS:protocol adapter error
TNS-00517: Lost contact
HPUX Error: 32: Broken pipe
05-FEB-2009 09:12:40 * (CONNECT_DATA=(CID=(PROGRAM=)(HOST=__jdbc__)(USER=))(SERVICE_NAME=rtst10g)) * (ADDRESS=(PROTOCOL=tcp)(HOST=172.30.0.13)(PORT=53349)) * establish * rtst10g * 0
05-FEB-2009 09:12:40 * (CONNECT_DATA=(SID=trial10g)(CID=(PROGRAM=rtsora@hpux1571)(HOST=hpux1571)(USER=fnsonlw4))) * (ADDRESS=(PROTOCOL=tcp)(HOST=172.30.0.12)(PORT=59918)) * establish * trial10g * 0
05-FEB-2009 09:12:40 * (CONNECT_DATA=(CID=(PROGRAM=)(HOST=__jdbc__)(USER=))(SERVICE_NAME=rtst10g)) * (ADDRESS=(PROTOCOL=tcp)(HOST=172.30.0.13)(PORT=53385)) * establish * rtst10g * 0
05-FEB-2009 09:12:40 * (CONNECT_DATA=(CID=(PROGRAM=)(HOST=__jdbc__)(USER=))(SERVICE_NAME=rtst10g)) * (ADDRESS=(PROTOCOL=tcp)(HOST=172.30.0.13)(PORT=53359)) * establish * rtst10g * 0

[Steven E. Protter]

Shalom,

I have seen this before on Oracle based systems. There are two areas to look at.

1) New OS patch requirements that Oracle has added since release.

2) Actual Oracle patches.

Both require a software contract with Oracle http://metalink.oracle.com

You will find these symptoms in Oracles site.

SEP

[Ankit Grover]

Thanks SEP!

I dont have direct access to Metalink.
I will reply with the details once I get the required details from our support team.

[Frank de Vries]

I had a simimar problem,
and used a workaround by adding a parameter to our sqlnet.ora

TCP.VALIDNODE_CHECKING = NO

But to be true, we installed some patches as well.

[Ankit Grover]

For the time being, we have added the following into listener file:

SUBSCRIBE_FOR_NODE_DOWN_EVENT_=OFF

The things are looking fine. Will observe it for few days.