Urgent: Workflow tasks not moving further due to multiple errors

meabhishek
meabhishek
edited May 14, 2012 in Documentum #1

Hi,

The workflow tasks are not moving further when the application is loaded with 30-36 users. In the repository log there are errors like "DM_WORKFLOW_E_PROCESS_AUTO_TASK", "DM_SESSION_E_CANT_MAKE_TEMP_CONNECTION".
The repository log has been attached with the SR for your reference.

Below is the Environment Details:
CS Version: 6.6
CS OS: Windows 2008 (64 bit)
Database: MS SQL 2005
Application Server: apache-tomcat-5.5.28
Web Browser: IE 7

Client Application: TaskSpace 6.6
Client Java Version: 1.5 Update 22

I'd hugely appreciate if someone can help me with the issue.

Thanks in advance for your replies.

Thanks,
Abhishek Pandey

Answers

  • Imran_Yusubov
    Imran_Yusubov
    edited August 8, 2011 #2

    Hi,

    1) I think this is

    Fri Aug 05 14:52:59 2011 781000 [DM_SESSION_E_CANT_MAKE_TEMP_CONNECTION]error:  "Could not establish an internal, temporary database connection (database_name = DM_hlikk_sqe_perf_docbase, docbase_owner = hlikk_sqe_perf, connection_string = HLIKKCONT01); database system error was:  STATE=42000, CODE=4064, MSG=[Microsoft][ODBC SQL Server Driver][SQL Server]Cannot open user default database. Login failed.

    not a workflow error.Have you tried reatarting your docbase it seems as if the database user is locked or the database is not reachable.As you can see the error clearly says the problem Login failed.Try to restart your docbase and see if it can connect to the database.If it can't connect to the database you will not able to start it.

    2) Regarding to this problem.

    Fri Aug 05 14:47:55 2011 562000 [DM_WORKFLOW_E_PROCESS_AUTO_TASK]error:  "Workflow Agent failed to process task 4a0030398002f063 of workflow 4d003039800034f6. The task is using method 'BPSIntegration'. Check the Java Method Server log for errors."

    That indicates an error in an auto activity.Please check if your Java Method Server is up and running.Then also  check the method server log , you can find the error details in there.Please attach it to the discussion too.

  • meabhishek
    meabhishek
    edited August 8, 2011 #3

    Thanks for the reply, Imran.

    The resolution to the first issue is correct, the database was not up.

    On the second error, the JMS was up while we got this error.

    To let you know, we are doing performance testing using automated scripts. The script accepts the tasks and forwards it to next actors. Can this problem be related with the Agent Worker Thread being non-responsive?

    Your thoughts please.

    Thanks,

    Abhishek Pandey

  • Imran_Yusubov
    Imran_Yusubov
    edited August 8, 2011 #4

    Can this problem be related with the Agent Worker Thread being non-responsive? I am not sure on that , are there anything in your java method server log?

    I have seen this error many times  while development of custom workflow methods.Then we have found out that the method is wrong or it is incorrectly configured so on.After correcting the problems with the methods the error had gone.So the method server log can let us identify the problem.

    Fri Aug 05 14:47:55 2011 562000 [DM_WORKFLOW_E_PROCESS_AUTO_TASK]error:  "Workflow Agent failed to process task 4a0030398002f063 of workflow 4d003039800034f6. The task is using method 'BPSIntegration'. Check the Java Method Server log for errors."

  • sagar.singh
    sagar.singh
    edited August 9, 2011 #5 
    meabhishek wrote:Thanks for the reply, Imran.The resolution to the first issue is correct, the database was not up.On the second error, the JMS was up while we got this error.To let you know, we are doing performance testing using automated scripts. The script accepts the tasks and forwards it to next actors. Can this problem be related with the Agent Worker Thread being non-responsive?Your thoughts please.Thanks,Abhishek Pandey

    Its definitely error due to auto-activities not getting executed by JMS.

    Can you please post the JMS error log for more information. Moreover one try you can make that stop and restart your JMS and look for the JMS server log file. If it says that port 9080, or 9090 is alreay used by some other resource then you can try to stop all the java.exe process from taskmanager and restart the JMS again.

    Try it,

  • meabhishek
    meabhishek
    edited August 24, 2011 #6

    Thanks all for the replies.

    The issue got resolved by increasing the number of WF Agent Worker thread.

    Thanks,

    Abhishek

  • klnschandra
    klnschandra
    edited May 14, 2012 #7

    Hi Abhishek,

    I am also facing the same issue.. Please provide me the information regarding increasing the number of WF Agent worker threads??

    Please find the environment details:

    CS - 6.7, Weblogic 10.3.3.0 Database- Oracle -11g

  • Vamsirk.masanam
    Vamsirk.masanam
    #8

    Hi all, can you please let me know hot to increase thewf agent threads on CS 7.2

Categories