IPA Reset process error status?

 6 Replies
 2 Subscribed to this topic
 52 Subscribed to this forum
Sort:
Author
Messages
mikeP
Veteran Member
Posts: 151
Veteran Member

    I've been trying to find a way to reset a process error status when a normal and expected condition occurs that IPA treats as an error.  The current issue is testing if a file exists.

    The File Access node can do that, but if it detects the file does not exist, it treats it as an error condition.  In this case, the file is probably present in 15% of the runs, and it is a normal condition for it not to exist.  If it's not found, the process does some housekeeping and terminates normally, but IPA flags the work unit as having errored, even if several subsequent nodes that include error handling do not encounter an error.

    I also tried a System Command node with a Windows command file that uses IF NOT EXIST to test the file, but the only way to pass that fact back to the process is via errorlevel which, as might be expected, sets an error condition in the process.

    I looked into forcing it with an Assign node, but system variables are not available for assignment.

    Right now, the only downside to this behavior is that errors appear in the work unit log and the work unit is displayed in red in the administrator, but it's not hard to envision a time when some kind of automated error test process might be developed to check for work unit errors, in which case this would be a false positive. 

    Any ideas on how one might reset the error condition?

    Woozy
    Veteran Member
    Posts: 709
    Veteran Member
      Hi mikeP - In the File Access Node, are you using the "Check file exists" mode, and did you check the "Ignore Error If File Does Not Exist" checkbox?
      Kelly Meade
      J. R. Simplot Company
      Boise, ID
      mikeP
      Veteran Member
      Posts: 151
      Veteran Member

        I don't see that checkbox, is it on the General tab of Properties? 

         

        Maybe it's in a later version of IPD...I'm on 10.0.6.12..6247 until we upgrade our Landmark server.

        Woozy
        Veteran Member
        Posts: 709
        Veteran Member

          That could be the problem.  Bummer.  I'm using 10.1.1.19.4412.  Here's a screenshot:

          Kelly Meade
          J. R. Simplot Company
          Boise, ID
          Terry P
          Veteran Member
          Posts: 234
          Veteran Member
            I think it's a bug in IPA 10.0 since in PFI there was a similar box called "Stop process on error". It is there in IPA 10.1 like you said - but that version of Designer is not backwards compatible with our version of Landmark we're on.
            Tim Cochrane
            Veteran Member
            Posts: 154
            Veteran Member
              mikeP - so you are wanting Completed flows to show with absolutely NO errors, even though there might be an error in the work unit...valid or not. Unfortunately that's not how it works, and you just have to figure out how to handle each error...whether within the flow or manual intervention. When i see a workunit with an error i look to see which node errored, then i either ignore it (since it's expected), or handle the error and re-run the work unit.

              I've seen some flows that do "Check file exists", then if it doesnt (which will always create a red error) then it will create the file and do the normal processing. On the "no error" leg the work unit deletes the file and stops. When i see this error, i know it's expected and i do nothing with it.
              Tim Cochrane - Principal LM/IPA Consultant
              JimY
              Veteran Member
              Posts: 510
              Veteran Member
                If Perl is available on the server you could use a syscmd node to run a Perl script to check if the file is there and write the result out to a file (Yes/No). Then you could use a file access node to read the file and pass the value on to a branch node to determine what your next step would be. Having said that I would leave the error in because if there is supposed to be a file then you wouldn't know until something else failed further down. You might also be able to configure your automated error test to ignore the error under certain conditions such as time of day, week, month,etc. My two cents worth.