Heartbeat program

[johnnyt]

Am new to Nodelink and in the DSCLink days (years) I had the following heartbeat monitoring program that I'm trying to convert:

Check Comms - DSC Copy - [ID 03AA][Parent 0016][Not Enabled]

If
        $sZDSC.Heartbeat is not 0
 
Then
        $sZDSC.CommsDown  = 0
        Wait  32 seconds
        $sZDSC.Heartbeat  = 0
 
Else
        Wait  5 minutes and 1 second
        $sZDSC.CommsDown  = 1

 

However I can't change the Heartbeat "device" in the NodeLink way of doing things. When I choose the DSC device (where the heartbeat shows up) it only offer options to query or trigger alarm. 

 

What does a heartbeat monitoring program look like in the new world order?

 

[paulbates]

Here is how I updated mine. While it looks unusual, here's what happens:

• The heartbeat should change every 30 seconds
• The program will re-execute every 30 seconds when it
  • either 1, or -1
• When the above happens, it never falls through to the error message
• If it does fall through, than something happened and there is a problem

Paul

[johnnyt]

Thanks, Paul. 

I used your approach, which made a lot of sense as usual, but have had a couple of false positives in just 3 hours of use. This happened even after changing the wait time to 95 secs, which would cover over 3 missed heartbeats.

I am getting some DSC connection errors/timeouts being reported in NodeLink but when it happened it didn't seem to be a 90+ sec issue.

2019-05-25 16:17:39 - DSC Error: Data Arrival Init - A connection attempt failed because the connected party did not properly respond after a period of time, or established connection failed because connected host has failed to respond [dsc1]
2019-05-25 16:18:20 - DSC TCP: No poll response, attempting to reconnect to alarm [dsc1]
2019-05-25 16:18:20 - DSC: Reconnecting To Server [dsc1]
2019-05-25 16:18:20 - DSC Error: Data Arrival Init - An existing connection was forcibly closed by the remote host [dsc1]
2019-05-25 16:18:50 - DSC: Reconnecting To Server [dsc1]
 

I can go to 2-3 or more minutes if needed since I'm unlikely to be able to deal with the issue immediately when it happens anyway but it doesn't seem to be an issue of the wait not being long enough as it is. That said I don't know what the errors really mean - could they be the kind of errors that take more than a few secs to recover from? Maybe duplicate occurrences are not logged?

What's your experience?

[paulbates]

Hi Johnny 

I forgot that since I posted that program, I had updated the time out to several minutes. I have 2 venstar stats and a RainMachine on one nodelink. I rarely have false positives and only from venstar but can simulate an error and pull a venstar off it’s mount and make its message happen. 

A few things

1- You have a lot of executing programs. Lots of executing programs and tight looped programs could cause the ISY to miss messages. Increasing the timeout is the only thing I can think of.

2- Also consider that if you use the venlink to nodelink migration method described in one of my previous posts, you’ll have numerous processes running side by side on your server wanting to hit the ISY;  and that might create contention leading to missed heartbeats. If you have dsc/venlink running, try temporarily shutting them down temporarily and see what happens.

Paul