Clovertech

‘unable to allocate fetch buffer’ TCL error

› Clovertech Forums › Read Only Archives › Cloverleaf › Cloverleaf › ‘unable to allocate fetch buffer’ TCL error

  • Creator
    Topic
  • December 18, 2012 at 4:07 pm #53418
    Gary Atkinson
    Participant

      I have a upoc thread that queries an sqlite database.  Every so often (probably once a week), the proc throws the tcl error–> ‘unable to allocate fetch buffer’.  Resending the errored message from the error message does not reproduce the error, so its not the message itself.  I am at a complete lose of why this *sometimes* happens 😳 Ideas??

      here is example error–>

      Code:


      [sms :sms :ERR /0:crisp_mck_hold:11/30/2012 20:16:10] Tcl error:
             msgId   = message473
             proc    = ‘tpsCrispDSQuery’
             args    = ”
             result  = ‘unable to allocate fetch buffer’
             errorInfo: ‘
      unable to allocate fetch buffer
         while executing
      “msgget $mh”
         (”run” arm line 3)
         invoked from within
      “switch -exact — $mode {
                     start {
                             #load $env(HCIROOT)/tcl/include/libtclsqlite3.so Sqlite3
                             package require libtclsqlite3 1.0
                     }
                     run {
                             ke…”
         (procedure “tpsCrispDSQuery” line 12)
         invoked from within
      “tpsCrispDSQuery {MSGID message473} {CONTEXT sms_ib_data} {ARGS {}} {MODE run} {VERSION 3.0}”‘

    • Creator
      Topic
    Viewing 4 reply threads
    • Author
      Replies
      • December 18, 2012 at 4:33 pm #77624
        Michael Hertel
        Participant

          {MSGID message473}

          Somehow you are not returning a dispList. (Memory leak)

          I’ve never seen messages much over double digits, let alone 473.

        • December 18, 2012 at 4:36 pm #77625
          Gary Atkinson
          Participant

            If a dispList was not being returned why does it not fail when I resend?

          • December 18, 2012 at 4:40 pm #77626
            Michael Hertel
            Participant

              Do you resend after you bounce the thread?

              That could clear the buffers although I would think the resend is handled differently by the engine.

              Does this error only happen to one message and continue normally without intervention or does it fail from that point forward with the rest of the messages until bounced?

              Further, you should be able to stop the interface and check the log to see if many messages are still bound to the interpreter.

            • December 18, 2012 at 5:01 pm #77627
              Gary Atkinson
              Participant

                No restarting done.  Seems to happen just so often and usually just one message.  One will error but others will be fine.

              • December 18, 2012 at 5:14 pm #77628
                Michael Hertel
                Participant

                  Hmmmm, well that’s weird.

              • Author
                Replies
              Viewing 4 reply threads
              • The forum ‘Cloverleaf’ is closed to new topics and replies.