Announcement

Collapse
No announcement yet.

Are missing devices detrimental?

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

    Are missing devices detrimental?

    If an event sends a command(s) to a device that is not present is there a trickle down impact?
    • For instance, if a device doesn't respond is there a pause on z-wave network? My assumption is not since the protocol seems very "UDP" like, but assuming there is some sort of collision avoidance. ,,,,so do other devices wait for it for "x" ms to respond?
    • Does HS4 do anything differently in program cycles?
    • Anything else?
    The reason I ask is because I'm rewriting an event that controls holiday lights. These all are on removable plugs and are only on the network when the lights are there. I already have routines that turn on/off lights morning/evening etc based on luminescence. Is it better to call another event from inside an event that can perform some sort of "if" check to not send the commands, or just as good to just add them to the existing events that run everyday of the year even when they are no on the network? Or, just create even MORE events that do nothing but the holiday lights that don't run when the plugs are not there.

    I'm sure the difference is probably small, but tons of things happen when these events run from HVAC changes, doors locking, inside/outside lights changing, timers kicking off, global variable comparison for temp/humidity changes, should the fireplace turn on, etc.

    My preference would certainly be to just keep it in the same event to avoid any further complexity (and I just want to know ).

    #2
    I use a group condition on all of my x-mas light. This is my preference - I do not like to have nodes that are not active on the system. I would consider plugging in the removable plugs in an out of the way duplex outlet.

    Click image for larger version

Name:	Screenshot from 2020-12-02 09-27-55.png
Views:	97
Size:	92.4 KB
ID:	1437696

    Comment


      #3
      Originally posted by ts1234 View Post
      If an event sends a command(s) to a device that is not present is there a trickle down impact?
      • For instance, if a device doesn't respond is there a pause on z-wave network? My assumption is not since the protocol seems very "UDP" like, but assuming there is some sort of collision avoidance. ,,,,so do other devices wait for it for "x" ms to respond?
      • Does HS4 do anything differently in program cycles?
      • Anything else?
      The reason I ask is because I'm rewriting an event that controls holiday lights. These all are on removable plugs and are only on the network when the lights are there. I already have routines that turn on/off lights morning/evening etc based on luminescence. Is it better to call another event from inside an event that can perform some sort of "if" check to not send the commands, or just as good to just add them to the existing events that run everyday of the year even when they are no on the network? Or, just create even MORE events that do nothing but the holiday lights that don't run when the plugs are not there.

      I'm sure the difference is probably small, but tons of things happen when these events run from HVAC changes, doors locking, inside/outside lights changing, timers kicking off, global variable comparison for temp/humidity changes, should the fireplace turn on, etc.

      My preference would certainly be to just keep it in the same event to avoid any further complexity (and I just want to know ).
      I doubt there is much impact on the event system but I'm pretty sure this causes problems with the Z-Wave network. Any failed communications seem to have a significant effect on the network performance. Z-Wave often appears to behave in a serial manner where waiting for responses slows everything down. In particular polling non responsive devices can bring the network to its knees. Similarly a faulty device spamming the network, or even a device sending too frequent power reports, interrupts other communications and introduces delays.

      So removing (not excluding) plugin devices but continuing to send commands to them, is likely to slow the network down.

      You also have to consider that these devices might be saved as neighbours and part of command routes for other devices on the network. I find that a Z-Wave network is a delicate thing which is easily knocked off kilter. Even if all your devices are Z-Wave plus, other devices might try routing through your removed devices before giving up and trying another route.

      I don't think Z-Wave is the best technology to use for modules that you don't want to be always powered. If you do though, I would recommend at least ensuring that you aren't trying to operate them when they aren't plugged in.

      Just my 2 cents

      Steve

      Comment


        #4
        I also use plug in modules for holiday lights, but in the off season I plug them into a power strip in the garage.

        Comment


          #5
          I just exclude the devices but save the event and disable it. This seems to work fine, but I recognize that routes may need to be re-optimized. But that would also be the case if you moved them to a power strip somewhere else in the house, right?

          Comment


            #6
            Thanks to everybody. Great input, I have 4 of Jasco ones that only 1 is "on net" for most of the year. I'd always felt like they'd last longer unplugged, but now that I think about what you guys have said maybe I'll plug them in. Tomgru's comment is probably most efficient and clean, but I was really hoping to make it so I could just plug them in knowing they'd just start working. I need to find me a z-wave sniffer. I can still remember the light going off working with Ethernet following communication flows through in a packet capture.

            Again...thanks everybody!

            Comment


              #7
              I use a handful of x-10 devices for this reason. Doesn't matter if they're plugged in or not if I don't need them. A wonderful advantage of this old technology...
              HS3PRO 3.0.0.500 as a Fire Daemon service, Windows 2016 Server Std Intel Core i5 PC HTPC Slim SFF 4GB, 120GB SSD drive, WLG800, RFXCom, TI103,NetCam, UltraNetcam3, BLBackup, CurrentCost 3P Rain8Net, MCsSprinker, HSTouch, Ademco Security plugin/AD2USB, JowiHue, various Oregon Scientific temp/humidity sensors, Z-Net, Zsmoke, Aeron Labs micro switches, Amazon Echo Dots, WS+, WD+ ... on and on.

              Comment


                #8
                At one time when upgrading my automation technology did try UPB, ZWave and Zigbee modules then went back to using X10 that I used to use 30 years ago.

                Today using X10 modules for Christmas decorations and my LED landscaping lighting.

                Outside utilize Black and Decker Freewire modules and inside I use appliance modules. All are set to one house code.

                I am entertaining trying modded WiFi switches outside. (Modded = > Tasmota or Espurna).
                - Pete

                Auto mator
                Homeseer 3 Pro - 3.0.0.548 (Linux) - Ubuntu 18.04/W7e 64 bit Intel Haswell CPU 16Gb- Mono 6.12.X - HSTouch on Intel tabletop tablets
                Homeseer Zee2 (Lite) - 3.0.0.548 (Linux) - Ubuntu 18.04/W7e - CherryTrail x5-Z8350 BeeLink 4Gb BT3 Pro - Mono 6.12.X
                HS4 Pro - V4.1.11.0 - Ubuntu 20.01/VB W7e 64 bit Intel Kaby Lake CPU - 32Gb - Mono
                6.10.0.104
                HS4 Lite -

                X10, UPB, Zigbee, ZWave and Wifi MQTT automation-Tasmota-Espurna. OmniPro 2, Russound zoned audio, Smartthings hub, Hubitat Hub, and Home Assistant

                Comment


                  #9
                  I've got a mix of "holiday" devices that only get plugged in once a year. I've got Jasco/GE outdoor receptacle z-wave plug-ins and now IKEA $10 Tradfri indoor plug-in Zigbee receptacles (best bargain ever). I've never had any communication issues or disruptions by simply plugging them in after a long absence. It really comes down to the stability/strength of your network mesh. I've carefully grown/planned all my home automation networks so coverage is never an issue.

                  Comment


                    #10
                    langenet / Pete / TC1 - I think I'll get mine all plugged in over the Holiday's and then see if I see anything different when I take them out while leaving them in events. After all, while they have not been in events, they have been devices not plugged in until now for quite a while. Besides, it would give me something to blame every time things didn't work right and there was no real smoking gun .

                    TC1 - I'm leaning a little your way that if they are gone, but nothing in my mesh cares from a hop/routing perspective, if that really has any serious/noticeable impact. I do want to find out how the zwave protocol addresses devices that do not respond. I'm guessing at things above what I'd call layer 2 in the OSI model don't care. I just don't know things like when a communication goes out from a device and no response comes back is there a waiting period for that response. My assumption is that the next packet just goes out as soon as there is a first gap in the traffic it can fit.

                    Comment


                      #11
                      Originally posted by ts1234 View Post
                      langenet / Pete / TC1 - I think I'll get mine all plugged in over the Holiday's and then see if I see anything different when I take them out while leaving them in events. After all, while they have not been in events, they have been devices not plugged in until now for quite a while. Besides, it would give me something to blame every time things didn't work right and there was no real smoking gun .

                      TC1 - I'm leaning a little your way that if they are gone, but nothing in my mesh cares from a hop/routing perspective, if that really has any serious/noticeable impact. I do want to find out how the zwave protocol addresses devices that do not respond. I'm guessing at things above what I'd call layer 2 in the OSI model don't care. I just don't know things like when a communication goes out from a device and no response comes back is there a waiting period for that response. My assumption is that the next packet just goes out as soon as there is a first gap in the traffic it can fit.
                      This is always a good reference to explain the operational workings of the network and resiliency for the Z-wave mesh:

                      https://www.vesternet.com/pages/unde...-nodes-devices

                      As noted, up to three different routes will be tried. Not noted in that link is that the network is always rebuilding the routing table organically as it changes. Many times people mess up the resiliency of their network by forcing routes.

                      Comment


                        #12
                        TC1 - Thanks! I probably have 10 other 100K z-wave devices with 15 feet of any of these plugs when in use. I'm guessing everybody can found a route when they are removed as you state. Since both the HS log and this article clearly state/show there really are ACK's (strangely I've read more than one thing that said they were not), and I have always staggered the start times of all of my reoccurring events so they don't overlap when running, I think I will just move any devices that get unplugged to the last ones in the event so no other step is waiting on them...maybe even pad them a bit with a "wait" and just see if I have any notable issues.

                        Comment


                          #13
                          Or possibly put the transient devices into separate events that only fire on a seasonal/time of year basis? This would be a good use for Group Event conditions (reference the HS3 Event Clinic forum for more info on this).

                          Comment

                          Working...
                          X