Forum Discussion

NAl's avatar
NAl
Contributor
4 years ago
Solved

Pipeline Status Notification

Hi there,

I know that we can receive notifications for Pipelines that run on Scheduled Tasks for a number of statuses such as started, stopped, failed and completed.

How can we do the same for a pipeline that fails or stops whilst running on a Consumer Snap where the Message Count is -1?

  • koryknick's avatar
    koryknick
    4 years ago

    @NAI - you can set up a scheduled task to run on an interval - however long you want your longest downtime. Then simply set the task so it doesn’t run if an instance is already running. That way you get notified if the job stops for any reason and it will automatically restart on the next interval.

    You may also want to look into Ultra Tasks (premium feature).

7 Replies

  • cstewart's avatar
    cstewart
    Former Employee
    4 years ago

    Run it as a task. That functionality is associated with a task.

    • NAl's avatar
      NAl
      Contributor
      4 years ago

      Hello,
      We can’t use a task for this specific pipeline as it’s constantly consuming messages from a queue.
      Is there an alternative means of doing this? We had an instance where the Snaplex lost connection which it prompted the pipeline to stop and the only way we discovered was through internal comms…

  • koryknick's avatar
    koryknick
    Employee
    4 years ago

    @NAI - you can set up a scheduled task to run on an interval - however long you want your longest downtime. Then simply set the task so it doesn’t run if an instance is already running. That way you get notified if the job stops for any reason and it will automatically restart on the next interval.

    You may also want to look into Ultra Tasks (premium feature).

    • NAl's avatar
      NAl
      Contributor
      4 years ago

      Hi Kory,
      Thanks for your response. This looks like a viable workaround.
      I wanted to check with you first whether the message below on the old version of article, How to Schedule a Task, is still valid?

      “A three month limit is placed on Do not start a new execution if one is already active . A second instance of that pipeline will be able to be executed at that time.”

      • cstewart's avatar
        cstewart
        Former Employee
        4 years ago

        The logic behind the 3-month limitation comes from the fact that we do a quarterly release, and the nodes on which the pipelines run should be restarting at least that frequently…