I have a 4 story stairwell in my house with secondary landings half way between each floor. I've automated the lights and I am using battery powered Zigbee motion sensors and I've also placed a Zigbee repeater (outlet) in the stairwell. I'm using 2 sensors per floor and they are grouped together as (for example) in a motion zone. This gives me 3 motion zones (1-2, 2-3, 3-4) and I'm having sporadic problems with response. Normally these zones are very responsive and work well but occasionally they just.don't.work. This problem had gotten so bad that I implemented a nightly (3AM) reboot of my hubitat.
That made a huge difference, but they still occasionally just don't work.
The only Zigbee I am using is motion sensors and outlet modules (mostly as repeaters.) The switches controlling the stairwell lights are all Z-wave. I have observed when the lights should turn on but don't that the Zigbee motion sensors detect me (based on the light flashing) but the Hub doesn't seem to get the notification (based on the logs.) At the same time, I can turn the stairwell lights on and off via the Hubitat app, so I know that Z-wave is solid to the lights.
All of my Zigbee devices show up as installed and there are no missing or extra ones. When everything is running well, they all respond very quickly. I don't have any 900mhz devices (that I am aware of in the house.)
Any advice on tracking this down this issue would be appreciated.
All I can think of is maybe if the individual timeout on one or more of them has already been triggered by a movement that affected only one of the pair, then when you stand in the zone that would have been seen by both, only the other one triggered. Or have you ruled that out specifically by seeing the indicator lights on both of each pair going together?
Where I have the motion sensors placed, it’s hard to see them both at the same time. One is placed to catch motion heading down from the top of a given floor to the intermediate landing and one is placed to catch motion coming up from below headed to the same intermediate landing. They work great most of the time, they just slow down occasionally, usually later in the day (since I implemented the 3AM daily reboot)
And is that the only rule or signal you have that seems to lag. Maybe you could do with the HubInformation community driver by thebearmay to see if there is some condition bothering your hub, if as you say it only slows down after being up for some time
Also as these are zigbee devices, is your zigbee on a channel that might get busy at certain times of the day for instance if a neighbor has their router on an interfering wifi channel. Or maybe even a microwave oven? Or don't they do that any more.
You shouldn't have to do a daily reboot. There is something wrong if you are rebooting that often.
Have you gone to the Logs tab, and over to App Stats, Device Stats and see if something is consuming excessive hub time. You will probably want to go to the Device/Hub Stats settings at the top of those pages, and select all items to get a more complete picture of what is happening. You can post screenshots if you aren't sure how to interpret those readings.
While you are in the Logs tab, also take a look at Scheduled Jobs and see if there is something that has excessive items. I don't know the exact number of what is excessive, but most of the time I only see 2-3 things scheduled for any one rule at a time. If you see dozens or hundreds of scheduled items, you should try to figure out why.
I suspect that whatever is slowing down Motion Zone is also making you have to reboot. Probably if you fix the root cause, everything else will run smoother.
This started happening to me after the hub update to 2.3.1. In the previous version my motion lighting was very quick to activate (also using Zigbee motion sensors with Philips Hue lights). However after updating to 2.3.1 all my motion activated lights rules slowed down very noticeably.
A rollback to the previous hub version fixes this however and the snappiness comes back.
Under 2.3.1, Over the course of a day from a fresh reboot, my hub free memory starts at around 450000 but then slowly goes down to less than 200,000. This has also necessitated a daily reboot for me to try and get the response times up under 2.3.1.
Regarding the response time, I haven't seen any difference between the older versions and 2.3.1. I don't think there were any changes to ML or Zones in that version that should have caused slowdowns, although there were a couple bugfixes in ML app.
Did you look at the App/Device stats I mention above? Anything wrong there?
I haven’t tried to problem solve it yet, other a rollback to confirm that it wasn’t happening on the previous version.
However I suspect that it may have something to do with groups and scenes which I believe were updated this version. But I haven’t done anything to test this hypothesis so it may not be the case. I use RM for my motion lighting rules and most active a scene rather than simply lights on or off.
I’ve posted by device and app settings from the logs below (not sure if that sheds light on anything). I’ve also checked the scheduled events and nothing there is showing any excessive scheduled events for a particular device or app.
To me, that App percentage of busy time looks high.
Those states are pretty high. Is there any reason you need to store hundreds upon hundreds of L360 states? Could you trim those back a bit?
And I would say that there is something going on with Hue app, and L360 app, that usage look very excessive compared to what I am used to seeing, but admittedly I don't have these exact apps on my Hubitat.
Same with "Study Lights with Keyboard", what is that, and why is it's usage so high up? Are you polling something or doing excessive refreshes?
Now I'm curious as to what's going on here, too. @darren3220 , do you mind sending me your hub's id in a PM? I can take a look at the engineering logs. Maybe there's something that can be tweaked in platform.
Checking my device list, I have 362 devices made up of a combination of virtual and physical. I’m not sure how to generate a list to work out which is which. But it looks to be approx as follow:
80 to 90 or so Hue Lights (spread over 2 bridges)
25 Kasa Smart switches (running old version of the community Kasa App pre cloud integration)
7 Wemo Smart switches
2 Harmony Smart Hubs
Broadlink Hub and an IR Blaster
Fronius Solar System.
Also out iPhones (4) connect via wifi as presence sensors.
Not many here. Just two hue hubs, and a Rachio Sprinkler
The rest are Zigbee, virtual, and Zwave.
Looking at my app list, 8 I think. My app list is below:
It may be I have this set incorrectly but can’t see where. However I have all my devices across the hub set to only store 11 events including the devices from the Life360 States App.
I have an app running in RM to detect presence in the Study so as to turn on and off the lights. The triggers are a motion sensor and a Samsung Multipurpose Sensor. The Samsung sensor is stuck underneath the keyboard and detects when the keyboard is being used as “active” so the rule knows not to turn off the lights even if I am sitting still. Originally I did have the sensor set to record 3 axis as well which completely flooded the hub with events. So that is turned off and only records active/not active. Looking at the event list it doesn’t seem to be triggering excessive events when in use.
I would have to put some thought into that Study Light rule and how to fix it (or defer to someone like @bertabcd1234) , but triggering on "change" for both of these is probably not the best thing to do. This rule is going to be running excessively due to those triggers. I almost think you need to split that up into two rules, one to decode the logic of presence, and then run this rule from that other logic rule.
I don't use L360, but I know others do, hopefully someone more familiar can give more specific advice. However what I do with many of these "chatty" things like weather and presence is to set the states as low as possible, in some cases 1 state. Unless you really need to go back in history, there isn't a reason to store so many. There used to be a lower limit of 11 before the hub would excessively do a cleanup process, but that was fixed a while back so you can safely reduce those states below 11.
Again I don't use Hue Bridge, but those look like they are storing excessive states. I am not sure who the expert with Hue is, but hopefully someone sees this and can make some suggestions about taming these a bit.