LinuxQuestions.org - OC4J instance restarts automatically after 10 seconds.

- AIX (https://www.linuxquestions.org/questions/aix-43/)

- - OC4J instance restarts automatically after 10 seconds. (https://www.linuxquestions.org/questions/aix-43/oc4j-instance-restarts-automatically-after-10-seconds-779383/)

OC4J instance restarts automatically after 10 seconds.

Hello Friends,IT System Administrators,

I am using Oracle Application Server 10.1.2 on IBM AIX 5.3 Platform. Since many days, my "OC4J" service restarts automatically in 10 seconds. By counting error logs in /ORACLE_HOME/opmn/logs/ipm.log,i have found below errors.
---------------------------------------------------------------
10/01/01 15:05:20 [4] Process Crashed: OC4J~WEBVAT~default_island~1 (864289932:626744) - Restarting
10/01/01 15:05:22 [4] Starting Process: OC4J~WEBVAT~default_island~1 (864289933:0)
10/01/01 15:05:29 [4] Process Alive: OC4J~WEBVAT~default_island~1 (864289933:729570)
10/01/01 15:28:29 [3] libopmnoc4j Could not connect to socket while pinging an OC4J process
10/01/01 15:28:29 [4] libopmnoc4j Process Ping Failed: OC4J~WEBVAT~default_island~1 (864289933:729570)
10/01/01 15:28:29 [4] Job 0 9969720 result:
ias_admin.gjvatwb1~OC4J~WEBVAT~default_island~864289933:729570
Status: Alive
Operation: internal (Connection refused)
ErrFile: /oracle/ora10gas/opmn/logs/OC4J~WEBPROBE~default_island~1
---------------------------------------------------------------

Kindly help me out to sort out this issue.

Nishith Vyas

Hi -

That's a very common error in AIX 5.3. It's just IBM's way of telling you that you should have purchased WebSphere instead of OC4J ;)

Just kidding ;)

ANYWAY -

It *IS* a relatively common error, with several potential root causes.

Google for "libopmnoc4j Could not connect to socket" and you'll find many hits. For example:

http://kr.forums.oracle.com/forums/t...hreadID=666949

Quote:

OK, there may be a reason for opmn to "think" your instance is dead. One reason is that the instance is so busy that it can't respond to opmn's requests.

For this case see : http://download-west.oracle.com/docs....htm#sthref946

--olaf

... and ...

Quote:

Please check the crrectness of the values in your servers conf file.
There might be an error in it.
Therefore you can have a look to the log file in opmn/logs/OC4J~OC4J...
if there is a configuration error -

for example:
Whitespace required
<repserver>.conf couldn't be parsed
at line: xxx : column: yyy
Expected name instead of "/"

When I adjusted my path for sourceDir and restarted the repserver the error was disappeared.

'Hope that helps .. PSM

replaced PING value from 4 to 5.

Hi "paulsm4",

I have just modified below line.
"log-file path="$ORACLE_HOME/opmn/logs/ipm.log" level="5""

Previously, it was aving value of "4" instead of "5". After that,i have restarted "opmnctl reload" & by measuring the ipm.log, i haven't found those messages i had received previously. Now, I am getting below messages.

Is it fine with the configuration?
------------------------------------------------------------------
10/01/03 11:22:33 [5] [libopmnoc4j] Ping succeeded for proc: 1692271770
10/01/03 11:22:44 [5] [libopmnoc4j] Received notification from proc: 1692271770
10/01/03 11:22:47 [5] [libopmnoc4j] Received notification from proc: 864289916
10/01/03 11:22:48 [5] [libopmnoc4j] Received notification from proc: 864289915
10/01/03 11:22:53 [5] [libopmnoc4j] Pinging proc: 864289916
10/01/03 11:22:53 [5] [libopmnoc4j] Ping succeeded for proc: 864289916
10/01/03 11:22:53 [5] [libopmnoc4j] Pinging proc: 864289915
10/01/03 11:22:53 [5] [libopmnoc4j] Pinging proc: 1692271770
10/01/03 11:22:53 [5] [libopmnoc4j] Ping succeeded for proc: 864289915
10/01/03 11:22:53 [5] [libopmnoc4j] Ping succeeded for proc: 1692271770
------------------------------------------------------------------

Guide me if the logs are saying different story. For Your Information that "website is running very fine" now. Though today is SUNDAY in india,i can check the website load on monday onwards.

I will revert back to you very soon.

Nishith Vyas

Hi -

It sounds like you were getting bona fide errors, and I don't think simply changing the log level from 4 to 5 would make them go away. I suspect that simply rebooting was what helped - and the problem might well start occurring again.

Please read through the example link I cited, and the links it cited. Please do your own Google search as I suggested, and see what else you find might be interesting. By all means, contact Oracle support (if your organization has it).

Good luck .. PSM