Announcement

Collapse
No announcement yet.

option to monitor powered devices too.

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    option to monitor powered devices too.

    Hi Steve.

    That's a great plugin. Thank you. AFAIK this plugin is only for battery devices? What about powered devices. I had problems with that too. Loosing connectivity to host, hardware failure.... Would be nice to have an alert option for those devices too. Any suggestions.

    Thank you

    Br,
    Dali



    #2
    Originally posted by Kdiamond View Post
    Hi Steve.
    That's a great plugin. Thank you. AFAIK this plugin is only for battery devices? What about powered devices. I had problems with that too. Loosing connectivity to host, hardware failure.... Would be nice to have an alert option for those devices too. Any suggestions.
    You are correct it is specifically for battery devices.

    I have had requests in the past to extend the pi for powered devices and, whilst I haven't ruled it out, I haven't really seen the need, at least on my system. There are so many different types of devices in HS3 that setting up monitoring would be complex and require a lot of input from the user and I don't want to cloud the central purpose of the pi which is focused on the specific issues of battery devices.
    If I were to include powered devices it would be need to be a different methodology to the battery devices. For example you wouldn't want a monitoring device created for every powered device.

    Steve

    Comment


      #3
      Why not just pool the powered devices on interval and alert if it doesn't respond. Wouldn't that be enough? Way more simple than battery devices that comes alive only at wake up period,

      Originally posted by SteveMSJ View Post
      For example you wouldn't want a monitoring device created for every powered device.
      Why not. I for example have half devices powered, half battery. I see no problem in that.

      Br,
      Dali

      Comment


        #4
        Originally posted by Kdiamond View Post
        Why not just pool the powered devices on interval and alert if it doesn't respond. Wouldn't that be enough? Way more simple than battery devices that comes alive only at wake up period,
        It's not that simple. You can poll Z-Wave devices because the Z-Wave pi exposes a function to do that but you can't poll other technologies unless the pi that owns them provides a method to do it. Even with Z-Wave, polling has a very heavy overhead. It takes a couple of seconds for a successful poll to return a result and an unsuccessful poll can take several seconds and it freezes the Z-Wave interface while it waits. Polling potentially hundreds of devices to check whether they are alive could bring your network to it's knees. In fact even polling a few devices sequentially causes problems. You don't have to look far through the forums to see the delays people have caused by polling. So, you have to be very careful what you do with a plug-in. SDJ-Health doesn't poll non-listening Z-Wave devices it monitors the wake-ups that they announce themselves so that doesn't add any overhead to the system. There are some listening battery devices which are always awake and it polls them but it staggers the polls out over several hours, one at a time so the impact is negligible.

        Originally posted by SteveMSJ View Post
        For example you wouldn't want a monitoring device created for every powered device.
        Originally posted by Kdiamond View Post
        Why not. I for example have half devices powered, half battery. I see no problem in that.
        Some people have hundreds of devices, so doubling the number of devices could be crippling. The point I was making though was that a different methodology would be required for checking powered devices. For battery devices creating monitoring devices make sense because I am storing information on battery levels, discharge rates, etc as well as optional configuration overrides. That probably wouldn't be necessary with powered devices.

        Having said all that, I do have ideas for adding some level of checking of powered devices which I will be experimenting with soon.

        Steve

        Comment


          #5
          Ahh, so too much poling is not good. Everyday I learn something new

          Maybe we can set expected update interval for each device and alert when it's silent for longer? Or pool only if device is silent longer than expected to check on it. That would eliminate too many pooling.

          Br,
          Dali


          Comment


            #6
            Originally posted by Kdiamond View Post
            Maybe we can set expected update interval for each device and alert when it's silent for longer?
            What I have done in this case is set up an event (A) that is triggered when the device is updated. When event A runs, it removes a delayed event (B) and schedules B to run in 12 hours. If event A runs in 2 hours, then the 12 hour event B will be removed (having 10 hours left) and set back to a 12 hour delay.

            Clearly event B can be anything HS can do, such as poll a device. For example:

            Event A:
            IF my-device just had its value set or changed.
            THEN Remove Delayed Execution of Event B
            THEN Run Event B After waiting 12h, 0m, 0s


            Event B:
            IF This event is MANUALLY triggered
            THEN Send an Email to me@somewhere with the subject my-device last update is too old Message: my-device last update is too old
            THEN Poll B for status.

            Len


            HomeSeer Version: HS3 Pro Edition 3.0.0.435
            Linux version: Linux homeseer Ubuntu 16.04 x86_64
            Number of Devices: 633
            Number of Events: 773

            Enabled Plug-Ins
            2.0.54.0: BLBackup
            2.0.40.0: BLLAN
            3.0.0.48: EasyTrigger
            30.0.0.36: RFXCOM
            3.0.6.2: SDJ-Health
            3.0.0.87: weatherXML
            3.0.1.190: Z-Wave

            Comment


              #7
              Originally posted by Kdiamond View Post
              Maybe we can set expected update interval for each device and alert when it's silent for longer? Or pool only if device is silent longer than expected to check on it. That would eliminate too many pooling.
              That's the kind of thing I am experimenting with but, as I don't see much benefit in monitoring powered devices, my intention is only to add it if I can keep the overhead on HS minimal. I personally (touch wood) don't have any problems with powered devices so introducing extra processing for no perceived gain is counterproductive. I'm wary of feature creep in a plug-in which is something I can be guilty of myself.

              By the way if you have just one or two powered devices that are problematical for some reason and you want to add then to your battery device monitoring in SDJ-Health you can actually do it with a bit of a work round. Basically what you do is add a virtual device which is the dummy battery for the powered device. You set the Device Type (string) to Battery and then group it as a child to the root device of the physical powered device you want to monitor, using Jon00's Grouping Script. In SDJ-Health you can then select this device to be monitored by Activity Checking and the pi will monitor all the child devices associated with that physical device. I actually use this for an old X10 motion sensor that I still use which doesn't have any battery reporting but does have a day/night light sensor which triggers twice a day. If you want to do this I can give you more details of how to achieve it.

              Alternatively, for one or two devices you could use a method such as Iveatch suggested above instead of trying to add them to SDJ-Health.

              Steve

              Comment


                #8
                It happened again today when testing devices. One powered NODE did not respond. It was powered plug (NEO COOLCAM NAS-WR01ZE Z-wave Plus Smart Power Plug EU) I did found out accidentally using Z-seer+. I don't know for how long, but it's dead. That's why we need some sort of monitoring and alerting of powered devices too.

                Br,
                Dali

                Comment


                  #9
                  Originally posted by Kdiamond View Post
                  It happened again today when testing devices. One powered NODE did not respond. It was powered plug (NEO COOLCAM NAS-WR01ZE Z-wave Plus Smart Power Plug EU) I did found out accidentally using Z-seer+. I don't know for how long, but it's dead. That's why we need some sort of monitoring and alerting of powered devices too.
                  I am working on an upgrade to SDJ-Health to check general devices in addition to battery devices but it isn't ready for testing yet. I'm concentrating on Z-Wave devices initially and have a basic experimental version running. When I have tidied it up a bit I'll post a beta version for you to test, as you seem to have a system that needs it.

                  Steve

                  Comment


                    #10
                    That's just perfect Steve.

                    I'll be happy to test it!

                    Let me know when you have it ready.

                    Thank you

                    Br,
                    Dali

                    Comment


                      #11
                      Originally posted by Kdiamond View Post
                      That's just perfect Steve.

                      I'll be happy to test it!

                      Let me know when you have it ready.

                      Thank you

                      Br,
                      Dali
                      Beta version for playing around with in new thread of this forum if you are still interested.

                      Steve

                      Comment


                        #12
                        Thank you Steve.

                        How does it work exactly?

                        What happens when I press RUN. Where do I get to see results or alerts? SDJ-Health Root?

                        Thank you
                        Br,
                        Dali


                        Comment


                          #13
                          Originally posted by Kdiamond View Post
                          Thank you Steve.

                          How does it work exactly?

                          What happens when I press RUN. Where do I get to see results or alerts? SDJ-Health Root?

                          Thank you
                          Br,
                          Dali


                          Did you read the guide attached to the new sticky thread?
                          Steve

                          P.S. Thread is at top 'General Health Checking'
                          Guide is attached to the first post General Health Checking Proof of Concept.pdf

                          Comment


                            #14
                            No I did not. Went directly to beta downloads

                            Will do, thx!

                            Br,
                            Dali

                            Comment


                              #15
                              Originally posted by Kdiamond View Post
                              No I did not. Went directly to beta downloads
                              Sorry for that, I should have been clearer.

                              Post any comments or questions in the new General Health Checking thread.

                              Steve

                              Comment

                              Working...
                              X