AppWorx, Dollar Universe and Sysload Community

  • Private Community
 View Only

  • 1.  Service IO Too Many Records for execution

    Posted Aug 12, 2019 10:34 AM
    Hello,

    Since the last online reorg wa have this message on our product_log from a Windows Server 2012 R2 Node on 6.10.1 :

    Service IO Too Many Records for execution

    More than this message we have many TCP TIME_WAIT connection from Local host to hostname on PORT 1700 :

    TCP 127.0.0.1:65473 FRGBMDUSAPW1:10700 TIME_WAIT
    TCP 127.0.0.1:65474 FRGBMDUSAPW1:10700 TIME_WAIT
    TCP 127.0.0.1:65475 FRGBMDUSAPW1:10700 TIME_WAIT
    TCP 127.0.0.1:65476 FRGBMDUSAPW1:10700 TIME_WAIT
    TCP 127.0.0.1:65477 FRGBMDUSAPW1:10700 TIME_WAIT
    TCP 127.0.0.1:65478 FRGBMDUSAPW1:10700 TIME_WAIT
    TCP 127.0.0.1:65480 FRGBMDUSAPW1:10700 TIME_WAIT

    Our product_log register more than 60 messages par seconds.

    And lots of execution are aborted by node without error message on product_log.

    Thank's for your help.

    ------------------------------
    Jean-Paul Gauger
    ------------------------------


  • 2.  RE: Service IO Too Many Records for execution
    Best Answer

    Posted Aug 13, 2019 04:20 AM
    The problem is resolve wich this procedure :

    The errors appear to originate due to an issue with the Launcher data files.
    I reinitialized the Launcher data files and the errors stopped being logged.
    Before restoring to the previous version, could you please perform the following procedure to see if it resolves your issue:
    ~~~~~~~~~~
    Warnings concerning this procedure :
    -THE IMPACT OF THIS MANIPULATION WILL BE THE DELETION OF THE PREVIOUS LAUNCHES.
    -THE LAUNCHES GENERATED BY A SCHEDULED TASK MUST BE RECREATED AFTER THE ENGINES ARE STARTED.
    -THE MANUAL LAUNCHES OR THE LAUNCHES GENERATED BY AN UXORDRE WILL HAVE TO BE MANUALLY RECREATED.
    Procedure :
    1. Copy / Paste the jobs in status "Launch Wait / Event Wait" from UVC - Job Runs or via the command "uxlst fla full" to know what was supposed to be launched next.
    2. Shutdown the Dollar Universe and check that all process are down
    3. Backup the launcher data files of the concerned area (u_fmsb60.dta, u_fmtp60.dta, u_fmsp60.dta)
    4. Load the Dollar Universe environment
    5. Reinitialize the launcher data files of the concerned area by executing the following commands from the bin folder
    Example for the area EXP (X):
     - uxrazfic u_fmsp60 X
     - uxrazfic u_fmsb60 X
     - uxrazfic u_fmtp60 X
    6. Reorganize the data files and then start the node again, using the following commands:
     - unireorg
     - unistart
    7. If tasks with "Start Date" in the past are not recalculated, launch the following two commands:
    uxupd tsk tsk=*
    uxupd tsk tsk=* ses=*

    Original Message