Odd arming failure behavior

I have HSM set to arm when mode changes to night with a 30 second delay. Night mode is triggered by my wife turning on her white noise machine which works great. Night mode also turns off the bedroom lights. I set up HSM to turn the bedroom lights back on in the case of an arming failure. This morning, the bedroom lights went on at 7:00 a.m. when a rule triggered day mode, which also disarms HSM. I don't have any rule set up to turn the bedroom lights on in response to day mode. I checked the status of sensors and found that one of my door sensors needs some recalibration. It was indicating open when the door was actually closed. That leads me to believe that the bedroom lights going on happened due to an HSM arming failure. But why would it have happened this morning instead of happening 30 seconds after mode changed to night?

Can you show how you have that set up? How are you arming HSM after a 30 second delay after going into night mode?

I didn't do anything fancy. Configuring arming and disarming based on mode changes is a standard feature, as are delayed arming and alerts for arming failures. I chose the 30 second delay because changing to night mode turns the bedroom lights off. If I didn't have a delay in there, the regular off action and the special on action would be in a race condition. With the delay, the behavior I would expect when arming fails is for the lights to go off and then come back on after 30 seconds. I have subsequently added a notification to my phone as an additional alert for an arming failure so I can be sure that is what is happening if it happens again.

No, they wouldn't. Because night mode if triggering the system to turn off your lights and arm the system to night mode. Then when night mode runs, it would sense that there is an arming fault and the lights would go on. However, I wanted to see how you had HSM set up. If you don't wish to share it, that's okay, but that will severely limit the help you are going to be able to get.

I'm not trying to hide anything. I described what I did, which was just menu picks. This isn't one of those cases of several interacting rules. There is no fancy code to show. Does this screen shot help? Note that the notify action was not there yet when the incident happened this morning.

Maybe I am not using the term race condition as you understand it. My point was that if the lights go off and then go right back on milliseconds later, I might think they never went off. If the lights go off and then back on 30 second later, it would be obvious that two separate events happened.

The problem is you are not understanding the sequence of events. The arming failure is going to happen when you ARM NIGHT. NOT after the arming delay. So, you have the exact same "race condition" you thought you were preventing. Having the arming failure after the delay would be completely useless. Once the system begins an arming count-down you can't stop it. You have to wait for it to finish and then disarm that system. So, having the arming failure then would be of little use. Give it a shot...and if you actually showed a full screenshot of your HSM setup, I'd be able to tell you how. But you still haven't shown a screenshot. This is still only a partial.

This debate we are having about the 30 second delay is not getting at the point of my question. I triggered night mode about 10:30 pm. That should have triggered arming HSM at about 10:30.30 pm. If arming failed, I would have expected the arming failure action (the lights turning on) to happen shortly after 10:30.30 pm. The lights went on at 7 am the next morning when day mode was triggered, which would trigger HSM to disarm. The 8 1/2 hour delay is what I am curious about - not the 30 second delay.

Perhaps I am misunderstanding the sequence. Here is my understanding of what I think the sequence of events should be:

  • 10:30.00 - My rule that triggers night mode changes the mode from Evening to Night.
  • 10:30.01 - My rule that turns out the lights based on this mode transition runs and turns off the lights.
  • 10:30.01 - Based on my HSM setup, the change to night mode causes HSM to wait 30 seconds and then attempt to arm.
  • 10:30.30 - HSM attempts to arm. If one of the contact sensors that would trigger an alarm is already open, then arming fails and HSM performs the disarm failure action.

Your statement that "the arming failure happens after you ARM NIGHT - not after the arming delay" doesn't make sense to me. According to my understanding, I don't ARM NIGHT until after the arming delay. In the HSM documentation, the example given for wanting a delay is someone using a button in their house to trigger ARMED-AWAY. The delay gives them time to open the door so they can leave without causing an alarm event. That example wouldn't make sense of HSM tried to arm before the delay had expired.

This is incorrect. It attempts to arm with a delay. The arming failure will occur here.