HS3 suddenly fails to send commands--HELP, HS is dead - Solved!

[KLisiak]

I am currently running HS3 (3.0.0.96) on Win 8.1 with a Z-Troller and USB adapter and 75 nodes. This setup has worked fine since I converted to HS3 and have had no problems in 9 months and have not made any changes during that time. Suddenly, I can't send any commands--polling, optimization or connectivity test all fail--though connectivity results are not quite stone dead. However, status changes are still being received, so communication is currently one-way only. I have local controllers which also seem to work, controlling local nodes. I also checked each switch by walking around with the Z-Troller and all respond with their node number locally. I changed batteries in the Z-Troller given previous commentary about low battery issues and rebooted the computer. No help.

I see that I am many revs behind on HS3, but am loathe to change anything on a system that suddenly stopped functioning. This failure hits right out of the blue like dead hardware.

Any ideas?

[cerjzc]

My first guess would be that something has started spamming commands as I have seen this several times. Sometimes I just needed to power cycle them, but other times it was an indication that something was about to die or had already stopped responding and needed to be replace. In my case every time it was a plug in module of some sort normally a lamp module although once it was an appliance module. If possible I would try power cycling all devices that might be plugin and worst case a "house reboot" might be needed. Once I had that not work and found that I had to locate the module that was causing the problem and remove it. That was not a fun weekend.

Good luck,
Jeff

[S-F]

Yeah. See my signature?

You may need to cut power to every Z-Wave device you have individually and test each time to determine which device is flooding your Z-Wave network and bringing it to its knees.

[TomTom]

My thinking is z wave communication bottleneck. Usually due to failed or failing( works sometime) device, too much polling, and/or too many aeon energy device reporting too much data back to controller.


Sent from my iPhone

[KLisiak]

So far, no success

Of the 75 nodes, only about 10 can be unplugged. Trying that produced no improvement. Starting in daylight, I'll take the batteries out of another 50. In your joint experience, do you see immediate improvement when the failing device is unpowered or do you need to restart HS3?

FYI--I tried the "audit" function and got mixed results. In instance could it "send a frame" successfully, but about half those I tried responded that they were able to send a "Query" and included the entire neighbor list of that node and related info. Those that failed in the query stated that they used a default reply stored in the PC. The status of devices on the Home page is accurate, but I would assess it a bit slower in catching up to reality than normal.

No optimization command has successfully started, after trying several nodes. Meanwhile, most timed events are working most of the time and the same is true of local controllers which are able to control their target devices most of the time. Does this fit the model of a failing node or should I suspect the Z-Troller or another controller? Is there a harmless way to restart the Z-Troller without losing any of the stored info?

Thanks for the support!

Ken

[randy]

Of the 75 nodes, only about 10 can be unplugged. Trying that produced no improvement. Starting in daylight, I'll take the batteries out of another 50. In your joint experience, do you see immediate improvement when the failing device is unpowered or do you need to restart HS3?

FYI--I tried the "audit" function and got mixed results. In instance could it "send a frame" successfully, but about half those I tried responded that they were able to send a "Query" and included the entire neighbor list of that node and related info. Those that failed in the query stated that they used a default reply stored in the PC. The status of devices on the Home page is accurate, but I would assess it a bit slower in catching up to reality than normal.

No optimization command has successfully started, after trying several nodes. Meanwhile, most timed events are working most of the time and the same is true of local controllers which are able to control their target devices most of the time. Does this fit the model of a failing node or should I suspect the Z-Troller or another controller? Is there a harmless way to restart the Z-Troller without losing any of the stored info?

Thanks for the support!

Ken

Three observations/suggestions

1. In my experience it is rarely a battery operated device that brings down a network
2. Power cycle the Z-Troller and restart your HomeSeer server
3. If you're still having problems kill as many circuits as possible until only HomeSeer and a device or two are powered. If it starts working bring up circuits one at a time and see if anything locks the Z-Wave network
   • Sometimes a device will cure itself when it is power cycled
   • Other times the defective device will still hose the network when it is powered up again

[KLisiak]

Z-Troller power cycle

By power cycle, I assume you mean disconnect and pull the batteries for 10 seconds or so? Do I need to do anything special when I restore power to the Z-Troller?
Ken

[randy]

By power cycle, I assume you mean disconnect and pull the batteries for 10 seconds or so? Do I need to do anything special when I restore power to the Z-Troller?
Ken

On the Z-Troller, remove the batteries and disconnect the power supply for at least 30 seconds, then connect the power supply to it. It will restart just fine and will not lose your network, unless the Z-Troller has a problem. They were pretty bulletproof so I doubt it has a problem. To be on the safe side, you can back up the interface before you do anything. There were many people who found a Z-Troller to lock up when the batteries were weak and the unit was powered by the wall wart. I didn't keep batteries in mine for that reason. If, after restarting the HomeSeer and power cycling the Z-Troller, you probably should leave the batteries out. At the very least put fresh ones in.

[KLisiak]

Success!!

Restarting the Z-Troller cured all problems!! Thanks to all who replied. Weak Z-Troller batteries seem to be the primary problem. Interestingly, the batteries were strong enough to do the Node check around the house, but too weak to avoid the lockup when dual powered. Go figure. Leaving the batteries out is great advice!
Thanks again,
Ken

[WfromL]

This incident illustrates perfectly the huge reliability problem with any z-wave setup: one device malfunctioning takes the whole network down. And finding out what exactly was the problem takes a long time, if it is possible at all.
I reluctantly moved to Z-wave some months ago because of the lack of support for X15 since HS3 arrived. I had three complete breakdowns of my system in one month, all due to z-wave and none of them with a clearly identifiable cause! I have been using X10 for 20 years, never did my system break down completely. Very occasionally one device failed, and then it is immediately recognizalbe which one. All 30-odd other devices remain functioning.
Sure, Z-wave has a lot more possibilities, but honestly, most of the time I only need an on/off switch, the rest is "luxury" which I gladly give up in exchange for reliability.
I am often abroad for a few weeks, HS3 is then in control of everything. A complete system breakdown in an empty house is a disaster.
So I am throwing out z-wave and returning to good old reliable - if simplistic -X10.

[KLisiak]

Reliability of HS3 and Z-wave

I concur on the concern with leaving HS3 "in charge" when away. This is the first time the cause was on the z-wave side. More common is that we either lose power or MegaSofty does something that reboots my PC. Does anyone have a suggestion on how to get a Win 8.1 (or Win 10 for that matter) PC to reboot directly into HS3? I had this working with the old WinXP setup and use a UPS to buffer short power outages, but that computer died.

Is Win 10 compatible with HS3 and is it any more friendly?

[peterpc]

This incident illustrates perfectly the huge reliability problem with any z-wave setup: one device malfunctioning takes the whole network down. And finding out what exactly was the problem takes a long time, if it is possible at all.
I reluctantly moved to Z-wave some months ago because of the lack of support for X15 since HS3 arrived. I had three complete breakdowns of my system in one month, all due to z-wave and none of them with a clearly identifiable cause! I have been using X10 for 20 years, never did my system break down completely. Very occasionally one device failed, and then it is immediately recognizalbe which one. All 30-odd other devices remain functioning.
Sure, Z-wave has a lot more possibilities, but honestly, most of the time I only need an on/off switch, the rest is "luxury" which I gladly give up in exchange for reliability.
I am often abroad for a few weeks, HS3 is then in control of everything. A complete system breakdown in an empty house is a disaster.
So I am throwing out z-wave and returning to good old reliable - if simplistic -X10.

I have good experience with Insteon.
And use the Insteon app as a fall back in case my PC 'hangs'.
(have an Insteon switch between my computer and can switch it with the Insteon app)

[randy]

Restarting the Z-Troller cured all problems!! Thanks to all who replied. Weak Z-Troller batteries seem to be the primary problem. Interestingly, the batteries were strong enough to do the Node check around the house, but too weak to avoid the lockup when dual powered. Go figure. Leaving the batteries out is great advice!
Thanks again,
Ken

I'm glad the controller reset cured the problem. I have seen it with Z-Troller and Z-Net interfaces and have exchanged messages with people who have had the same problem with USB controllers as well. It is not a very common occurrence - it happened twice with my Z-Troller and once with my Z-Net, over 2+ years. When you are having Z-Wave network problems, the controller is the first place to look. Leaving the batteries out of my Z-Troller ended problems with that device. I'm not sure what happened with my Z-Net, but since moving it to a stable power supply (via POE), it hasn't happened again.

When I wrote above

I If you're still having problems kill as many circuits as possible until only HomeSeer and a device or two are powered. If it starts working bring up circuits one at a time and see if anything locks the Z-Wave network

• Sometimes a device will cure itself when it is power cycled
• Other times the defective device will still hose the network when it is powered up again

I should have added - If it doesn't appear to be the controller and you decide to see if you have a bad device actively jamming the network, kill circuits at the breaker panel one at a time. Keep notes as you kill circuits so you can try to know which one "unlocks" the network. If a power cycle of a device clears the problem, it might be only a temporary cure. The device could start acting up again. If the problem comes back then you can kill the individual devices on that circuit one by one. Most switches and dimmers will have an air gap switch that will allow you to disconnect the device without having to remove it from the wall box.

When you are diagnosing a problem it pays to try to narrow your search as methodically as possible. Start with the controller, then circuits and finally devices - line powered devices before battery powered. Battery powered devices can become troublesome, but if they are killing the network, their batteries won't last very long.

[randy]

I concur on the concern with leaving HS3 "in charge" when away. This is the first time the cause was on the z-wave side. More common is that we either lose power or MegaSofty does something that reboots my PC. Does anyone have a suggestion on how to get a Win 8.1 (or Win 10 for that matter) PC to reboot directly into HS3? I had this working with the old WinXP setup and use a UPS to buffer short power outages, but that computer died.

Is Win 10 compatible with HS3 and is it any more friendly?

Starting with Windows 8, I used Startup Delayer from r2 Studios. I am now running Windows 10 and still using it. It can be used to delay the startup of HomeSeer, a very good idea if you are using Ethernet to serial (or USB) devices. It also pays to delay HomeSeer bu 30 seconds to let Windows finish most of its startup tasks. It has advanced options to allow you to start a program with the highest privileges. The application is free.

Windows 10 is no more or less "friendly" than Windows 8 (8.1), but it is back to being more desktop oriented by default. My HomeSeer server was equally reliable under Windows 7, 8.1 and Windows 10 - just be sure and defeat automatic Microsoft updating.

[randy]

This incident illustrates perfectly the huge reliability problem with any z-wave setup: one device malfunctioning takes the whole network down. And finding out what exactly was the problem takes a long time, if it is possible at all.
I reluctantly moved to Z-wave some months ago because of the lack of support for X15 since HS3 arrived. I had three complete breakdowns of my system in one month, all due to z-wave and none of them with a clearly identifiable cause! I have been using X10 for 20 years, never did my system break down completely. Very occasionally one device failed, and then it is immediately recognizalbe which one. All 30-odd other devices remain functioning.
Sure, Z-wave has a lot more possibilities, but honestly, most of the time I only need an on/off switch, the rest is "luxury" which I gladly give up in exchange for reliability.
I am often abroad for a few weeks, HS3 is then in control of everything. A complete system breakdown in an empty house is a disaster.
So I am throwing out z-wave and returning to good old reliable - if simplistic -X10.

Actually it doesn't illustrate a "huge" reliability problem, it does show that the technology can have problems. The OP's problem was probably caused by having old batteries in a Z-Troller, a problem that can only occur with a Z-Troller, because the rest do not have batteries.

From my perspective, Z-Wave has been very for quite a while. Insteon seems like it might be essentially more reliable, but Z-Wave still has a lot more choices in devices and controllers. It may be the tightly controlled product supply that keeps Insteon more reliable. I have seen many X-10 problems over the years due to power line noise, inadequate phase coupling, etc. While there is still X-10 available, if I was going with a power line communication topology, I would go with Insteon. We are fortunate that HomeSeer allows us to choose the technology interface we wish to use.

If I had it to do all over again knowing what I do right now, I would still choose Z-Wave.