Z-Wave Unresponsive / Delayed Response

My OpenHAB setup:

• Intel Xeon E3-1235 / unRAID VM / 2 core / 1G RAM
• Ubuntu 18.04.4 LTS / openjdk version "1.8.0_242"
• OpenHAB 2.5.3-1
• Aeotec Z-Stick (3 years old--I think Gen5)
• 28 Z-Wave devices (2 "ghosts")

Issue:

Sometimes when I turn on or off a Z-Wave device via HABPanel, OpenHAB Android client, or Echo + MyOpenHAB it does not respond immediately if at all. This does not happen all the time, but is rather frequent. And sometimes everything will work fine with near instant responses. This started happening after I made some changes over the last holiday season. Those changes were:

• Updated from my stable (no delays ever noticed) 2.4 setup using the Ubuntu packages provided. I did this mainly to be able to add a garage door opener that needed secure inclusion.
• Added a garage door controller. This had several challenges to get securely included, so I was trying different things. The final thing that got this working was changing the network key, (and I had to change more than just a couple bytes in it). I found this in a topic in this community.
• Changed to use MyOpenHAB instead of HUE emulation for Amazon Echo integration. This is because a firmware update to the Echos broke the HUE emulation. I'd rather keep as much as I can on premises, but since the automation is now a part of our lives I couldn't leave things broken for long.

Analysis:

I do not recall when the issue started and I made these changes in quick succession (I know, stagger changes…) so cannot say which thing was directly involved.

I have been searching this forum off and on since this started and have seen a few similar topics. However they either did not apply to me or did not seem to be resolved. Taking advise from one post I am creating a new topic as my issue may or may not be the same as someone else’s.

Last night I turned on Z-Wave debug logging and captured this:

The log begins with a fresh start of OpenHAB and letting it run pretty much untouched overnight, (I maybe turned on one thing through the Echo and a few manual light switch toggles). This morning I went through the house turning on and off most all switches through the Echo. Everything worked until the very end of the test when I started to see delays.

There is no specific device that seems to trigger issues. I think most everything defined has shown a delay in response at some point. Last night (before clearing the logs) things worked for a bit then like this morning started showing delays, but on a different device. Sometimes the device never seems to respond, other times it may respond after 30 seconds to a couple minutes. There is NO LOAD on the server when there are delays, (CPU usage is under 1% and there’s plenty of free RAM available). When one device stops responding it appears that other devices also show delays.

I would appreciate any help or ideas people have. I have done my best to research this, but things either go over my head or do not seem to apply.

You may have noticed in your forum search that there are some threads about folks who upgraded then experienced delays with zwave. Typically, delays are from one (or more) devices flooding the network with messages causing the delays. Most times it turns out to be a node which has low batteries or a so called ‘zombie’ node. If you have a Windoze PC you can use, try installing the PC controller app if you suspect a zombie node. You will have to search recent post on this forum but I’ll try to find link. Just general advise but hope it helps

Yes, I did see a few similar posts. It seemed like the recommendation was to use a “zniffer”. I do not have that currently. Or is there some other way to check for flooding without a modified hardware device? I do have an old laptop with Windows 10 on it, so that part is available.

PS: @Bruce_Osborne, thanks for adding the tag. Whenever I tried to submit this with the tag I got a 500 Internal Server Error message. It would only post for me with no tags.

You are welcome.
zwave tends to give the tag error but z-wave works for me.

Search for zwave PC controller software
On phone right now

I finally got around to looking at this more. I continue to have issues with ZWave commands being delayed sporadically.

I downloaded the PC controller software onto a laptop, but it seems like I would need to disconnect the Aeotec controller from my OpenHAB server and plug into the laptop. And I wasn’t sure what I was supposed to do to diagnose. There are a lot of options in the software.

So, I bought a UZB1 Z-Wave controller, flashed the Zniffer firmware and installed the Zniffer software. I put the laptop near the Aeotec and let it capture traffic. Again, I am not sure for what I am looking.

I do not have a lot of data at this point, but in the most recent case where there were delays I saw the following:

  • The LED on the Aeotec stick was “frozen” on a single color (instead of cycling through colors)
  • There were about 12 messages about “Assign SUC Return Route” on the screen, 6 saying “Routed:(1) ->31 - (33)” and another 6 saying "Routed:(1) -> 32 - (33). If 1, 31, 32 and 33 are Z-Wave nodes, these are plug-in smart switches that I have installed during the holidays to control lights, but are unplugged most of the rest of the time.
  • There were no messages for 4 minutes and during that time no Z-Wave commands went through, (and nothing was shown in the capture).

So, I have some tools now, but not really sure what to do next. Any tips on how to diagnose, or if the above suggest something please let me know.

Edit: I haven’t seen anything that looks like what I’d expect a message flood to look like. There is a small flurry of traffic when I turn on a light or something, but nothing that goes on for more than a second or two. And otherwise there is basically no traffic in the capture except for some bursts here and there.