Post Processing requests

[latifa]

Hi Erman,

submiting requests is finishing with warning messages. The OPP log showing this quind of error:

Exception in thread "OPPAQMON:43495" Exception in thread "ViewMonitor of J12345:10.182.2.190:[e-mail]" java.lang.OutOfMemoryError: Java heap space


What does it mean ?  What should I check ? How can correct this ?

Regards,
Latifa

[ErmanArslansOracleBlog]

It means Java heap size ( set with Xmx and Xms) is not enough.

Read these ->

http://ermanarslan.blogspot.com.tr/2014/09/ebs-122-output-post-processor.html

http://ermanarslan.blogspot.com.tr/2014/12/ebs-122-late-charges-report-xdodtexe.html

[latifa]

Hi Erman,

the log file errors doesn't looks like your log. Please find the logfile of my OPP.

I updated fnd_cp_services as following:

UPDATE fnd_cp_services
SET developer_parameters =
'J:oracle.apps.fnd.cp.gsf.GSMServiceController:-mx1024m'
WHERE service_id = (SELECT manager_type
FROM fnd_concurrent_queues
WHERE concurrent_queue_name = 'FNDCPOPP');


Is it enough ? How about the time out that I saw in some requests ? where to modify it ?

FNDOPP43495.txt



Regards,
Latifa

[ErmanArslansOracleBlog]

check first;

select DEVELOPER_PARAMETERS from FND_CP_SERVICES
where SERVICE_ID = (select MANAGER_TYPE from FND_CONCURRENT_QUEUES
where CONCURRENT_QUEUE_NAME = 'FNDCPOPP');

Then do the update above. (I wrote 1024 but you can make it 2048 as well)

update FND_CP_SERVICES
set DEVELOPER_PARAMETERS =
'J:oracle.apps.fnd.cp.gsf.GSMServiceController:-mx1024m'
where SERVICE_ID = (select MANAGER_TYPE from FND_CONCURRENT_QUEUES
where CONCURRENT_QUEUE_NAME = 'FNDCPOPP');
then restart concurrent managers.

Then, restar te conc managers and retry.

[latifa]

Hi Erman,

It wasn't enough and customer still have error messages and failing requests. Here are for some particular requests the recorded error messagees :

1. One or more post-processing actions failed. Consult the OPP service log for details.
CONC-POST-PROCESSING RESULTS
-- PUBLISH:
oracle.apps.fnd.cp.opp.PostProcessorException: oracle.apps.fnd.cp.opp.PostProcessorException: oracle.apps.xdo.XDOException


2. ****Warning!!! Due to high volume of data, got out of memory exception...***
****Please retry with scalable option or modify the Data template to run in scalable mode...***

3. I translate: Post treatment request failed with  the error message:
wait time for the out put processor  which should terminate this request is declared as passed  by the concurrent manager. Check if there are enough service processes for the Out Put Processor which are executing.


Regards,
Latifa

[ErmanArslansOracleBlog]

You can try setting the Xmx and Xms values in the concurrent's command line..

1.  Navigate to the System Administrator responsibility.
2.  Navigate to Concurrent - Program - Define.
3.  Query up the XDOTMGEN executable short name.
4.  In the Options field add a value such as -Xmx512m or -Xmx1024m or other relevant number to increase the heap size.
5.  Save the changes.
6.  Resubmit the request.

as for the warning -> "Warning!!! Due to high volume of data, got out of memory exception", send me the name of the concurrent program/programs

[latifa]

Hi Erman,

The program name is  XLAAARPT and PAXAARPT.

The failed request was executed with the first one.

Regards,
Latifa

[latifa]

Sorry, I confused. Here it is the programme name :

XDODTEXE.

Regards,
Latifa

[ErmanArslansOracleBlog]

Did you read this ?->
http://ermanarslan.blogspot.com.tr/2014/12/ebs-122-late-charges-report-xdodtexe.html

[latifa]

Hi Erman,

Thank you Erman for your reply and Sorry for the delay. Yes I read this note but and I didn't make all the changes described in it:

o FND_CP_SERVICES (2048 instead of 1024).
o Threads : from 5 to 10
o Timeout  : 120 to 20000 (s)
o Treatment Delay : from 300 to 20000 (s)

Regards,
Latifa

[ErmanArslansOracleBlog]

So, let's try Latifa..

[latifa]

Hi Erman,

the situation is more better than before, not much errors except one with the following error message in the FNDOPPxxxx.txt:

[11/24/17 4:39:07 PM] [OPPServiceThread0] Post-processing request 1342746.
[11/24/17 4:39:07 PM] [44534:RT1342746] Executing post-processing actions for request 1342746.
[11/24/17 4:39:07 PM] [44534:RT1342746] Starting XML Publisher post-processing action.
[11/24/17 4:39:07 PM] [44534:RT1342746]
[11/25/17 3:45:47 AM] [UNEXPECTED] [44534:RT1342746] java.lang.ThreadDeath

For other requests, they finish successfully but they took long time : 5 hours like that.

Others requests duration is about 45 mins --> do you think it is long time ?

Size of output files ~ 276M --> do you think that it's big size ?

Regards,
Latifa

[ErmanArslansOracleBlog]

The java.lang.ThreadDeath error indicates that the Output Post Processor has reached its processing timeout.

Let's try this ->

1)Increase the value of the Concurrent:OPP Reponse Timeout profile option to 300 secs. (if it is lower)
2)Login as SYSADMIN.
Responsibility: XML Publisher Administrator.
Function: Administration
Set the following properties:
Temporary Directory
Use XML Publisher's XSLT processor: True
Enable scalable feature of XSLT processor: True
Enable XSLT runtime optimization: True

3)Restart the Concurrent Managers

4)Retry

[latifa]

Hi Erman,

Concurrent:OPP Reponse Timeout profile option is set to 20000.

I set XSLT processor, scalability and runtime optimization.

Let see what it will come from these changes.

Regards,
Latifa

[latifa]

Hi Erman,

The request finished with following error message :

One or more post-processing actions failed. Consult the OPP service log for details.

CONC-POST-PROCESSING RESULTS

-- PUBLISH:
oracle.apps.fnd.cp.opp.PostProcessorException: oracle.apps.fnd.cp.util.RemoteFileException: java.net.ConnectException: Connection timed out

please find the most recent FNDOPPxxx.txt.
FNDOPPxxx.txt


Regards,
Latifa

[ErmanArslansOracleBlog]

Did you see the following?

Concurrent Processing - Potential Output Post Processor - OPP Failures Due to a Multi-Node Setup (Doc ID 563233.1)  ***Section : "3.1. Incorrect routing rules - java.net.ConnectException: Connection timed out"**

[latifa]

Hi Erman,

Thank you for the Doc ID, it's really the right one.

Our balancer balance beteew OHS servers via 8001. Hwo can we make loadbalancer communicate with GTS and the others ?

Regards,
Latifa

[ErmanArslansOracleBlog]

What do you mean by "GTS and others"?

What happens is : "The Output Post Processor attempts to transfer the XML data file by setting up an HTTP connection to the source node and this process fails at the load balancer."

So check your load balancer side..  routing rules...

[latifa]

Hi Erman,

Thank you for your reply.

I mean that how can I check that loadbalancer has a problem. I don't have to configure that (service is like this here). But I should ask them to configure it for me if I can justify that this is the problem.

For the person who is in charge of loadbalancer, he asked between what and what because actually loadbalancer is configured between OASs only (web).

Sorry I always forget and write in french: GTS means Concurrent managers servers CM

Others means I think OAS servers (http servers).

Regards,
Latifa

[ErmanArslansOracleBlog]

Basically, check your Load Balancer configuration. two way -- both from load balancer to apps nodes and from apps nodes to load balancer..