Help! Report R5642565 not printing full batch!

Shruthi Manas

Member
About once a week UBE R5642565 ECMD0810B ends on the Iseries, killing the JDE kernel, while processing a batch of EDI Invoices. When this happens the job failed to load all the batch invoices into the F470461. To finish processing the batch, our JDE/EDI technician resubmits the same R5642565 ECMD0810B job, selecting only those missing invoices. This second submission runs fine without issues.

Please let me know if anyone has faced a similar issue and how can I fix it?


TIA,
Shruthi
 
Shruthi,

Being that it's custom - we really don't have enough info to do much more than the guess work.

In the UBE's log (WSJ, Select the UBE, Row-Exit to Logging...) - are there any hints why it failed?

(db)
 
Dan is right... your UBE is a custom object so there isn't much the user community can do to help you without you providing much more details.

I would start with finding all the differences between R5642565 and R42565. With those differences, is there any custom c-BSFN's being used? If so, that would be a good starting point for verification of any memory leaks.

Hopefully you have a non-production environment that resembles your production environment in terms of EDI setup... if so, copy the failing batch of invoices to that environment and try running it there. Also, I'd try the same batch of invoices and run the R42565 to see if it fails also. The idea of this test being: if you can make R5642565 fail and the same batch of invoices run through R42565 runs successfully, your issue most likely resides in the differences.

Is there any similarities in terms of batch size when a batch fails? Are they always larger batches?
 
What does the log for R5642565 say?
Does the JDE/EDI technician submit it to the same server, or locally? Using the same user ID?
Can you reproduce the problem at all? In test/dev?

From your description, a "similar issue" could be almost anything, you need to supply some information to narrow it down a bit.
 
Sounds like a memory issue of sorts. CNC will be able to tell you why the job failed.

For starters: Make sure all your table io is wrapped in Open/Close where possible
 
Thank you all,

Will do all the checks from your inputs and keep you all updated.


Regards,
Shruthi
 
If your job is running early in morning or during off night hours in scheduler you also need to look into the backup time. A job log or jde log will give some useful info for the batch ask from CNC for the logs when job fails.
If you have CRP environment then you can process the same batch in CRP and check for any issues.

Chan
 
Here is the debug log

I'am unable to upload the entire log as its too big


**********************************************************************************
May 29 12:09:07.690136 - 832/5644 WRK:Starting jdeCallObject Entering jdeCloseDictionary
May 29 12:09:07.690137 - 832/5644 WRK:Starting jdeCallObject Exited jdeCloseDictionary with DDType 0
May 29 12:09:07.690138 - 832/5644 WRK:Starting jdeCallObject Return value is 0 for RetrieveVertexDatabaseConnection. (BSFNLevel = 1)
May 29 12:09:07.690139 - 832/5644 WRK:Call Object Entering AddObjectCache
May 29 12:09:07.690140 - 832/5644 WRK:Call Object Entering JDB_ValidateHUser (User 1C802D28)
May 29 12:09:07.690141 - 832/5644 WRK:Call Object Exiting JDB_ValidateHUser with Success (User 1C802D28)
May 29 12:09:07.690142 - 832/5644 WRK:Call Object no action properties to send, action property count = 0
May 29 12:09:07.690143 - 832/5644 WRK:Call Object no item properties to send, item property count = 0
May 29 12:09:07.690144 - 832/5644 WRK:Call Object Error count: Sent back - Server new #Error[0], #Warning[0]: Both Server&Client #Error[0] #Warning[0]
 
Can you run it for just one Line of a SO?
You should get a smaller log then.

Did it used to work and have you just changed something? If so compare the ER to PD and DV. See what's changed
 
Shruthi,

Can the logs be zipped / compressed ?

Regards,

Sanjeev


I'am unable to upload the entire log as its too big


**********************************************************************************
May 29 12:09:07.690136 - 832/5644 WRK:Starting jdeCallObject Entering jdeCloseDictionary
May 29 12:09:07.690137 - 832/5644 WRK:Starting jdeCallObject Exited jdeCloseDictionary with DDType 0
May 29 12:09:07.690138 - 832/5644 WRK:Starting jdeCallObject Return value is 0 for RetrieveVertexDatabaseConnection. (BSFNLevel = 1)
May 29 12:09:07.690139 - 832/5644 WRK:Call Object Entering AddObjectCache
May 29 12:09:07.690140 - 832/5644 WRK:Call Object Entering JDB_ValidateHUser (User 1C802D28)
May 29 12:09:07.690141 - 832/5644 WRK:Call Object Exiting JDB_ValidateHUser with Success (User 1C802D28)
May 29 12:09:07.690142 - 832/5644 WRK:Call Object no action properties to send, action property count = 0
May 29 12:09:07.690143 - 832/5644 WRK:Call Object no item properties to send, item property count = 0
May 29 12:09:07.690144 - 832/5644 WRK:Call Object Error count: Sent back - Server new #Error[0], #Warning[0]: Both Server&Client #Error[0] #Warning[0]
 
This looks like the concluding log. Can you look for any error. Go thru the Oracle document or Search for list to know how to read the log and try to see if you find any error in the log which can help you give the root cause of failure. Also did you check the jde.log on server during that time that can also give some useful system info.

Chan
 
Hi John,

I looked for failed and error in the log file and this is what i found, Is there any checks I can do with this?

I'am new to JDE and thanks to all of you and the forum for helping out

May 29 12:04:15.994000 - 832/716 FOREIGN_THREAD CBPMSocket::Connect(localhost, 9876) failed
May 29 12:07:41.451067 - 832/200 WRK:Init Remote Env Token Entering JDB_GetLastDBError
 
Back
Top