1 Reply Latest reply on Feb 12, 2018 4:37 PM by drodriguez

    Bringing value retrieval method attribute into an Event Frame template

    drodriguez

      Hey everyone,

       

      I'm trying to bring in what I call a "fault count" attribute into an event frame template. I plan to use it as a triggering condition for a notification rule.

      This is what it looks like on an example element:

      It does an event weighted Total of the Device Status Fault attribute over the past week.

      The parent attribute does exception handling and looks at boundary conditions.

      In the event frame template, I try to directly bring in the child attribute

      The parent attribute has the same configuration as it does in the element. It needs the Device Status Fault attribute to work, so I bring it into the event frame too.

      There's a lot going on here, but the central issue is bringing the child attribute in.

       

      This is one of the event frames from the same example element. I expect the child attribute to have a value in the 900's, but instead I'm getting 1

      One reason the might explain the value of 1 is that when the event frame attribute calls the element attribute, the element attribute somehow only covers the time range of the event frame, instead of the past week like it normally does.

      I think the sure shot solution here would be to use an analysis to write the Fault Count values to a tag, then pull that into the event frames. It would be nice to avoid running more analyses, and I'm curious as to why this doesn't work. Any thoughts?

        • Re: Bringing value retrieval method attribute into an Event Frame template
          drodriguez

          In case anyone is still wondering about this, I went ahead and added a line to my fault check analysis that writes the current fault count to a new attribute I created. The attribute writes to a tag with archiving disabled, since I don't need the history. As expected, the event frames bring in that value just fine. One thing to keep in mind is that because archiving is disabled, the event frame attribute value won't make sense when looking back at past event frames. For my purposes, this is acceptable.

           

          I probably should have lead with this, but the overall idea here is about notifications for interface device statuses. I have an analysis that will write a value of 1 to an output tag when the device status tag has a value that I consider a fault (and write 0 when back to normal). This is then used to trigger a notification. It works great for most interfaces, but I've found that some interfaces, from time to time, can rapidly fluctuate between a good and bad device status. This causes an excessive number of notifications to come in. Manually disabling and re-enabling notification rules for the large number of interfaces we monitor is out of the question, so the event frame attribute I created serves to automatically suppress the notification (via notification rule trigger conditions) when an interface has more than 20 device status faults over the past 7 days. And it's been working great!

           

          Hope this helps someone, and feel free to ask me questions.

          1 of 1 people found this helpful