Confused by lack of inactivity reporting

[RRR]

Steve,

I've just recently started to utilize your PI for battery and activity monitoring, although I've had it installed for a little while. Basically, finally sat down to read the documentation after noticing a failed device was not reporting in. That same device, Zooz Temp & Humidity Sensor, is still included in HS, yet I'm not getting any type of alerts for inactivity. I've setup the Events according to the documentation and I due receive emails for batteries drain and precent left.

I thought... maybe since this Device was already dead prior to me really setting up the PI correctly (assuming I have) that is why I'm not getting inactivity alerts. So then I pulled a battery from a Zooz Leak Detector sitting on my desk two days ago. It too has not alerted me to inactivity. I've included some screen snaphots to help explain this. I don't understand how this Leak Sensor, with no battery, has a "Last Good Check" from yesterday afternoon. Battery was pull on the 19th.

The third and fourth snapshot is from the Temp & Humidity Sensor. It's physically not in my house anymore and has been sent back for replacement.

I've also included the configuration, most of it, in the fifth snapshot.

What am I doing wrong?

[SteveMSJ]

Steve,

I've just recently started to utilize your PI for battery and activity monitoring, although I've had it installed for a little while. Basically, finally sat down to read the documentation after noticing a failed device was not reporting in. That same device, Zooz Temp & Humidity Sensor, is still included in HS, yet I'm not getting any type of alerts for inactivity. I've setup the Events according to the documentation and I due receive emails for batteries drain and precent left.

I thought... maybe since this Device was already dead prior to me really setting up the PI correctly (assuming I have) that is why I'm not getting inactivity alerts. So then I pulled a battery from a Zooz Leak Detector sitting on my desk two days ago. It too has not alerted me to inactivity. I've included some screen snaphots to help explain this. I don't understand how this Leak Sensor, with no battery, has a "Last Good Check" from yesterday afternoon. Battery was pull on the 19th.

The third and fourth snapshot is from the Temp & Humidity Sensor. It's physically not in my house anymore and has been sent back for replacement.

I've also included the configuration, most of it, in the fifth snapshot.

What am I doing wrong?

Hi RRR,

I'll try my best to figure out what is going on with your setup.

Firstly, is there a reason that you have deselected 'Monitor Log for Wake-ups'? For Z-Wave non-listening devices this is by far the best way of monitoring whether they are alive or dead. This method also has the benefit of the devices being picked up automatically by the plug-in and monitoring devices created. However, you can deselect it if you wish and for other types of devices the alternate monitoring methods are necessary.

One anomaly I can see is that the status reports in your small screenshot show 'Monitoring 16 devices' but on your 'Config' screen shot it shows 4 devices selected for monitoring by Polling and 15 devices selected for monitoring by Activity, i.e. 19 in total. Have you got the same devices selected in more than one list? How many monitoring child/features have you actually got, I presume 16?
It shouldn't actually be possible to select more than one monitoring method for a device, or at least if you do the plug-in will default to the method with the highest priority and remove the device from the other list. This doesn't happen immediately though but when a check is done. Have you selected any devices in either list recently, since the status report screenshot?

Looking at your Mud Room Water Leak Sensor, this would appear to be being monitored by Activity, based on the wording 'Last Good Check'.
In your Config you have the Period set to 24 hours. What that does is check at 24 hour intervals that there has been some activity, on any of the features of that device, within the previous 24 hours.
The Last Good Check is shown as 2/20/22 3:49 PM
The screenshot of the device shows the last change time as 2/19/22 9:02 AM
As you say there is something wrong there as at the time of the Last Good Check there hadn't been any activity in the previous 24 hours. I don't understand why that should happen and I've never seen it before.
Are there any hidden child/features for this devices?

Similarly there is something odd with your Gym Temperature & Humidity Sensor.

I'll do some testing in case there is some bug from a recent change in HS4 or the plug-in. Which version of HS4 and SDJ-Health are you running?

Of my 20 battery devices I have 5 monitored by Activity so, as a test, I have reduced the 'Period' down to 1 hour (normally I have it set to 12 hours, which is the default) and the OverRun to 1 minute.
One device immediately went to 'Silent' and I received a notification. That device is a doorbell, which doesn't normally show frequent activity in any of it's features except for the battery in the bell which updates at least every 12 hours. The plug-in correctly detected that there had been no activity in the previous hour so set the monitoring child to 'Silent' and raised the alert.

I also pulled the batteries from a Spirit radiator valve. These devices update regularly as they include a temperature reading amongst other features. It had updated just prior to pulling the batteries so I'll report back later on the results of this test.

Steve

[SteveMSJ]

I also pulled the batteries from a Spirit radiator valve. These devices update regularly as they include a temperature reading amongst other features. It had updated just prior to pulling the batteries so I'll report back later on the results of this test.

The plug-in correctly alerted me about the Spirit valve with removed batteries at the expected time.

Everything seems to be working correctly in my tests and I haven't had anybody else report issues.
I need to delve deeper to figure out what is happening or your system. If you could answer the questions in my first response I will investigate further.

Steve

[RRR]

Steve, let me see if I can address all of you questions and provide and update on my end as well.

Some good news, the Mud Room Leak Sensor has gone into Missed Activity Alert (although I don't receive an email for it - even after clicking Retrigger - I'll have to check on the event trigger if that is included). Screen shot #1. Why now thou...? When it should have trigger yesterday, since I pulled the battery on the 19th. Here's what I think..., I was browsing the forums trying to find answers for what I was seeing and read one of your previous comments about PollDevices should be enabled if you have z-wave locks, and I do. Even thou all of my z-wave devices were accounted for in Monitor Activity (15 of 16 enabled) I decided to enable PollDevices too. Very shortly later I posted this original thread on the forum. I'm wondering if enabling this level of check had anything to do with it. The leak sensor is not included in the 4 PollDevices thou. You would know better than me thou.

The Gym Temp and Hum. sensor still has not missed activity. Screen shot #2. Maybe it will this evening with the next 12 hour period lapses.?.?

Still leaves unexplained where those two sensors received Last Good Checks when they physically couldn't have. Difference of only using the one monitor? IDK, just trying to think of what it might have been.

When I read the documentation, I felt like I didn't have a need and shouldn't, leave the other monitoring features enabled if not really used. I thought you had mention wake-ups and heartbeats are something that aren't used much anymore. Plus being a long time, Systems Engineer I'm always looking for ways to minimize system resources/impact. Apparently, I misunderstood. My thought was if I can get by with just Monitor Activity then I wouldn't necessarily need the other three monitoring features.

There are a lot of questions here, let me address them inline, in Red...

---

One anomaly I can see is that the status reports in your small screenshot show 'Monitoring 16 devices' but on your 'Config' screen shot it shows 4 devices selected for monitoring by Polling and 15 devices selected for monitoring by Activity, i.e. 19 in total. Have you got the same devices selected in more than one list? Yes How many monitoring child/features have you actually got, I presume 16? Correct, 16 total individual devices with battery features (one doesn't actually have a battery - not sure why the feature is listed, so I disabled it under Monitor Activity). As I mentioned previously I had very recently enabled PollDevices monitoring feature. That then added 4 of the same devices.
It shouldn't actually be possible to select more than one monitoring method for a device, or at least if you do the plug-in will default to the method with the highest priority and remove the device from the other list. Its about 11 hours later and I can still select the same device in the two different monitors. This doesn't happen immediately though but when a check is done. Have you selected any devices in either list recently, since the status report screenshot? I think my previous comments address this as well.

Which version of HS4 recently upgraded to 4.2.8.0 and SDJ-Health 3.1.09 Beta are you running?

---

I hope I addressed all your questions. I'll keep an eye on the one Temp and Hum. sensor to see if it goes into alert as well. Back to work tomorrow so my responses will be delay, but let me know if I can provide any additional details/logs/screen shots.

Thanks Steve!

[SteveMSJ]

Steve, let me see if I can address all of you questions and provide and update on my end as well.

Some good news, the Mud Room Leak Sensor has gone into Missed Activity Alert (although I don't receive an email for it - even after clicking Retrigger - I'll have to check on the event trigger if that is included). Screen shot #1. Why now thou...? When it should have trigger yesterday, since I pulled the battery on the 19th. Here's what I think..., I was browsing the forums trying to find answers for what I was seeing and read one of your previous comments about PollDevices should be enabled if you have z-wave locks, and I do. Even thou all of my z-wave devices were accounted for in Monitor Activity (15 of 16 enabled) I decided to enable PollDevices too. Very shortly later I posted this original thread on the forum. I'm wondering if enabling this level of check had anything to do with it. The leak sensor is not included in the 4 PollDevices thou. You would know better than me thou.

The Gym Temp and Hum. sensor still has not missed activity. Screen shot #2. Maybe it will this evening with the next 12 hour period lapses.?.?

Still leaves unexplained where those two sensors received Last Good Checks when they physically couldn't have. Difference of only using the one monitor? IDK, just trying to think of what it might have been.

When I read the documentation, I felt like I didn't have a need and shouldn't, leave the other monitoring features enabled if not really used. I thought you had mention wake-ups and heartbeats are something that aren't used much anymore. Plus being a long time, Systems Engineer I'm always looking for ways to minimize system resources/impact. Apparently, I misunderstood. My thought was if I can get by with just Monitor Activity then I wouldn't necessarily need the other three monitoring features.

There are a lot of questions here, let me address them inline, in Red...

---

One anomaly I can see is that the status reports in your small screenshot show 'Monitoring 16 devices' but on your 'Config' screen shot it shows 4 devices selected for monitoring by Polling and 15 devices selected for monitoring by Activity, i.e. 19 in total. Have you got the same devices selected in more than one list? Yes How many monitoring child/features have you actually got, I presume 16? Correct, 16 total individual devices with battery features (one doesn't actually have a battery - not sure why the feature is listed, so I disabled it under Monitor Activity). As I mentioned previously I had very recently enabled PollDevices monitoring feature. That then added 4 of the same devices.
It shouldn't actually be possible to select more than one monitoring method for a device, or at least if you do the plug-in will default to the method with the highest priority and remove the device from the other list. Its about 11 hours later and I can still select the same device in the two different monitors. This doesn't happen immediately though but when a check is done. Have you selected any devices in either list recently, since the status report screenshot? I think my previous comments address this as well.

Which version of HS4 recently upgraded to 4.2.8.0 and SDJ-Health 3.1.09 Beta are you running?

---

I hope I addressed all your questions. I'll keep an eye on the one Temp and Hum. sensor to see if it goes into alert as well. Back to work tomorrow so my responses will be delay, but let me know if I can provide any additional details/logs/screen shots.

Thanks Steve!

Hi RRR,

This plug-in started out many years ago as a relatively simple tool which monitored wake-ups of non-listening Z-Wave battery devices (which most Z-Wave battery devices were at the time) and provided a way of being alerted in a centralised manner without needing to configure events, etc for individual devices. Over the years, in response to requests and suggestions from users, as well as my own requirements, it has developed into the complex plug-in you see today. Partly this is to provide methods of dealing with the huge variety of battery devices and technologies, and how they are represented in HS, but also due to the desire to make things as configurable as possible. I also added other features to the plug-in beyond battery devices. With this continued development and evolution the User Guide may not be as user friendly as it should be😕 That's also partly due to it being a free plug-in.

To figure out what has gone wrong with your setup would probably involve me requesting all sorts of screenshots and debug logs. Each monitoring device has it's own configuration page which, if required, can be used to override global configured parameters. So, there are lots of places to look when trying to figure out what has gone wrong.

I would suggest the best course of action in your case would be get everything back to defaults, effectively starting the monitoring from scratch. Whilst you could retain the parent device and adjust configurations so that you don't need to alter the event you have set up it would be easiest to start the plug-in setup from scratch.

The various monitoring methods for battery devices together with their advantages and disadvantages are explained in section 4.2.1. of the User Guide. The monitoring method is used to determine whether a devices is alive or dead. Once a monitoring child/feature has been created the battery level, rate and life monitoring are the same whichever monitoring method is used.

If you are already using the 'General Devices' and/or 'Log Monitoring' sections of the plug-in, then it might be best to retain the configurations but otherwise I would suggest you do the following:

• Disable SDJ-Health.
• Delete the SDJ-Health Parent and all Monitoring children/features. Best done using the Bulk Edit in HS4.
• Delete the 'SDJ-Health.ini' file from your HS installation 'C:\Program Files (x86)\HomesSeer HS4\Config\SDJ-Health.ini', or wherever you have it installed on your system.
• Make sure you have ‘Log Poll and Wake-up Messages’ checked in ‘Plugins>Z-Wave>Controller Management>Z-Wave Networks and Options’
• Enable SDJ-Health.
• Go to 'Plugins>SDJ-Health>Battery Devices', where all settings should now be at their defaults.
• Make sure that 'Monitor log for wake-ups' is selected, it should be as that is the default.
• Wait, say 24 hours, for all your non-listening battery devices to wake-up at least once.
• The plug-in will detect each one as it wakes-up and create a monitoring child. The status of the each child will initially be set to 'Waiting' until it wakes-up a second time whereby the plug-in then knows the wake-up interval for that particular device.
• If any of your non-listening Z-Wave devices aren't detected automatically check that they are working.
• Whilst you don't need to wait for the non-listening devices to be detected before proceeding, it is less likely to lead to confusion. Non-listening devices will appear in the list of devices available for Activity Checking until a wake-up has been detected and a monitoring child automatically created.
• If you have Z-Wave listening battery devices, typically locks, sirens, some radiator valves, they use Flirs or Beaming to essentially appear to the controller to be always awake. They don't wake-up at predefined intervals and therefore won't be detected automatically. These are best monitored using the Polling method. On the 'Plugins>SDJ-Health>Battery Devices' page select them in the list called 'Select devices that require polling'. In due course the plug-in will create monitoring children/features for these devices.
• There are a few non-listening Z-Wave devices that don't issue wake-ups, 1st generation Z-Smokes being an example. These can be monitored automatically from their Heartbeats (select that option on the 'Plugins>SDJ-Health>Battery Devices' page. The plug-in will then automatically create monitoring children/features for these and monitor their heartbeats. If you don't have any then leave this option deselected as the default.
• Z-Smokes significantly changed the way they work for the 2nd generation models so the Heartbeat method doesn't work. They can however, be monitored using 'Activity Checking'.
• Activity Checking is used for all other types of devices. It can be used for the above devices as well but it relies on devices updating on a regular basis so isn't necessarily suitable for all devices. Unchecking the 'Last Change Time Updates on Status Change Only' for appropriate child/features that update regularly but without the value changing can improve Activity Checking.
• If you have devices that need to be monitored using Activity Checking then select them from the 'Select devices that require Activity Checking'. If you have any battery devices that you wish to monitor and they don't appear in this list then it might be necessary to edit their 'Device Type String' so they are detected as Battery Devices. There are a few forum posts on this but feel free to ask how to do it.
• Note that some devices can be battery operated or powered. HomeSeer shows a battery child/feature for these devices even if they are included when powered. These will appear in the lists of battery devices but as long as you don't select them they won't be monitored.
• Once you have everything detected and working with the default settings create a suitable event or events as described in section 4.1.1. of the User Guide.
• After this you can adjust settings globally and/or individually to fine tune your system and devices.
• Whilst LogToDatabase is 'Off' by default I would suggest turning it 'On' as it improves the usefulness of the information the plug-in can show.

It's a while since I started from scratch so the above might not be complete so if anything is unclear or you have further questions please ask.

Let me know how you get on.

Steve

[RRR]

Hi Steve,

Thanks for the detailed steps to start a new. I'll give it a try this upcoming weekend.

I did want to pass along some good news on the Temp & Hum. Sensor. It did go into alert finally last night. Screen shot #1. As for why I didn't receive an email on Missed Wake-Up (0), turns out I didn't have that value included in the event range. I only had a range of Low Battery (1) - Battery Life Reached (3). I don't recall, this evening, if the documentation called this out, but I remember it showing an example of the range of 1 - 3.

And then some not so good news... Last night I added the battery back in the Mud Room Leak Sensor, it checked in immediately and the alert cleared in SDJ-Health. All good! Then just before retiring for the evening I decided to pull the battery again. Not 24 hours just yet, however the PI is reporting a Last Good Check of this afternoon, not possible. Screenshot #2. Last night I enabled all 4 monitors in PI as well. Wondering if there is a variable that is not cleared out/null before being reused - it's happened to me a time or two...

Anyways, I appreciate all of your help and I'll start over given time.

[SteveMSJ]

Hi Steve,

Thanks for the detailed steps to start a new. I'll give it a try this upcoming weekend.I
....

And then some not so good news... Last night I added the battery back in the Mud Room Leak Sensor, it checked in immediately and the alert cleared in SDJ-Health. All good! Then just before retiring for the evening I decided to pull the battery again. Not 24 hours just yet, however the PI is reporting a Last Good Check of this afternoon, not possible. Screenshot #2. Last night I enabled all 4 monitors in PI as well. Wondering if there is a variable that is not cleared out/null before being reused - it's happened to me a time or two...

I'm not sure of the timing but because you are using Activity Monitoring and you have the Period set to 24 hours, when it did the check at 10:13 PM there presumably had been activity in the previous 24 hours so it would show good at that time?
Setting a long period using Activity Monitoring will mean a long lag before reporting a failure. With your Setting of 24 hours it will be between 24 and 48 hours after the last activity before an alert. Setting a short period results in more false failures, depending on the particular device's reporting frequency. the default of 12 hours works fairly well. That is why Activity Monitoring is the lowest in the hierarchy of methods in the plug-in.

There isn't a variable to be cleared out. When you enable Wake-Up monitoring nothing will change until each device actually logs a wake-up. The wake-up period will then be adjusted for each device when it wakes-up a second time. With wake-up monitoring you have to give the plug-in time to detect and adjust to the wake-up sequence of each device, some of which may be 12 hours or more. Once Wake-up monitoring has configured itself it is very reliable but you need to be patient during the initialization period😊

The typical event in the guide triggers on the range of Missed WakeUp to Battery Life which is a range of 0 to 3, if talking in values. However you could have different events triggering for separate alerts of you want to get more complex. For example I use an event triggering on the change to Healthy to send me a pushover message when all failures have been cleared and everything is now good.

I think you will only confuse yourself by chopping and changing your setup. It will all make more sense if you start again and follow through the procedure I have suggested 😊

Steve