Skip to end of metadata
Go to start of metadata


  • Spools hang or get delayed during printing with status "Waiting for output formatter"



  • For this error message "Waiting for output formatter", there can be two possible reasons:
  1. Availability of spool processes.
  2. Missing commits in the application program that creates the spool.



  • Availability of Spool Processes:
    1. Increase the spool processes per server to 4 or possibly more.  See note 108799:
        • 108799 - How many spool work processes for each instance?
      1. You can increase the number of spool processes by setting parameter rdisp/wp_no_sp
      2. If you are printing many documents via Frontend printers like LOCL, you may also need to increase parameter rdisp/wp_no_spo_Fro_max.  The number of spool processes available for Frontend printing is set by this parameter
    2. Also check if the option "Process Requests Sequentially" is selected in the Output Attributes tab in SPAD for the output devices which have these problems
      1. When you send many output requests to an output device with 'Process Requests Sequentially' selected, then all of these prints are processed by only 1 spool work process (because this is the only way to guarantee that the requests will be processed in the right order)
      2. If you must use 'Process Requests Sequentially', then the best solution is to attach spools to previous spool requests in your application(s) instead of creating a new spool every time, as described in these notes:
        • 85318 - Appending documents to existing spool requests
        • 16410 - Attaching to existing spool requests
        • 65109 - Long delays when printing during overload
  • Missing commits in the application program:
    1. For the problem of missing commits in the application, refer to the note 529966:
        • 529966 - Long waiting times for print output (for example, with RFC)
    2. Spool work processes only begin their work when they are informed by a commit work from the application.  This can be done by an explicit commit work in the program or by an automatic commit work at the end of the program.  Until there is a commit work, a spool work process does not know about the new request.
    3. It may also help to append the spools for each job to one larger spool for applications that create many spools in a short time, instead of creating many smaller spools (as described in point 3 of note 65109 mentioned above)


  • No labels