Several zwave devices have gone unresponsive today

Stuck on or stuck off. This was before I took the latest upgrade but continues after it. WTF? These devices have been working for years, though I have noticed all z-wave is getting slower and slower and slower to respond and no amount of rebooting or restarting seems to help.

Losing confidence in you hubitat....

Awh, @bill. What have you done to troubleshoot this specifically? You can't just post an "i'm pissed off" post and not expect someone to ask what's happening. Do tell. I know you want to.

4 Likes

Well I was on my way out the door. Issues have been ongoing for several months. Devices disappearing from the network then re-appearing after restarting them or after z-wave repair. Sometimes some devices take up to 30-40 seconds to respond, other times they're fine.

This morning a device I've never had trouble with in the 3 years I've had it went un-responsive, then another. I fully restarted the hub, then installed the latest update and rebooted and repaired z-wave. Still un-responsive I thought when I wrote this message, but it turns out they were just taking 30-45 seconds to respond. All appears to be working again, except one lock that doesn't resond. Usually I can pull the battery and restart it and its seen again but not the case this time.

Steps I've taken:

Disabled all non-essential apps like z-wave poller
Turned off my 2.4ghz wifi
restarted the hub
rebuilt the z-wave mesh multiple times
I can't really think of anything else I can do

I'm baffled especially by the intermittent slowness. Sometimes a device responds in a few seconds, sometimes it takes 30-40 seconds. This problem has appeared slowly over the last couple of months. I don't quite understand how something digital can "degrade" like this and survive across reboots. Dunno if its at the hub processing level or at the z-wave or zigbee mesh level or what, and can't really think of a way to debug.

Sorry if you've posted in the past that you have tons of repeaters, but...do you have enough repeaters? Are they spaced adequately? How long are you waiting between Z-Wave Repairs?

I've no idea how many devices you have and how many are repeaters vs end devices, but rebuilding neighbor tables can take a long time. Clicking Z-Wave Repair a lot isn't going to do you any favors. I don't have a lot of Z-Wave devices, but it is often screwed up until the next day after a Z-Wave Repair is initiated I find.

Have you checked the live log for errors?

I have several of the old centralite iris z-wave/zigbee smart plugs and I have the z-wave repeater portion of them turned on... or I did. I suppose I should check they're still there....

yeah still there. There's one 3 ft from the device that was acting up this morning.

OK weird. I just checked my apps and somehow z-wave poller was still there, and watchdog. I just deleted both. I thought I'd disabled them but they appeared to still be operational.

I've never owned one of those, but my understanding from those that have is the Z-Wave portion is difficult to get working. It is possible that it's failed or turned itself off somehow, or factory reset somehow and the Z-Wave radio isn't working in them? It would only take one falling off to really screw up the mesh. Z-Wave doesn't rebuild the mesh by itself like Zigbee does. They try to do automatic rebuild of the neighbor tables with Explorer Frames in Z-Wave Plus, but it's not perfect and no where near as good at rebuilding as Zigbee is. And with those not being Z-Wave Plus, that's not even an option.

FYI. Just posted today. Enlightening info that may help.

1 Like

This is certainly interesting but it's all a bit ephemeral. I don't suppose anybody has any info on what versions of the centralite are affected, or driver version? Is there any way to test them or get more detail out of the z-wave mesh to see where they're being used?

I recall it being difficult to setup the z-wave portion of them, but generally I've been pleased with the centralite iris stuff. It's a shame they're gone. It was good stuff at a fair price, and available locally.

If anyone does, it would be @srwhite

You have low signal or interference.. Perhaps some of your repeaters aren't connected or didn't pair correctly (simply a guess).

Let's start here... Are they Centralite-branded plugs, or Lowe's Iris? Centralite never released a branded plug with Z-wave built-in, those were exclusive to Lowe's only. Look for the Z-Wave logo on the label. If it's missing, it's not a Z-Wave repeater.

Now, if they are the Lowe's branded ones, did you pair the Z-Wave radio too? It's a separate and somewhat frustrating process that is different than pairing the Zigbee side of the plug. I posted a driver for the Z-Wave side of those plugs that lets you interrogate and monitor the repeater side to make sure it's functioning. It's a potentially useful tool for troubleshooting mesh issues like that. The Z-Wave radios in these plugs for security reasons will only pair when in range of the hub, they do not support network-wide inclusion. After pairing these devices it's a good idea to run not 1 but 3 separate Z-Wave repairs, giving some time in between for each device to learn its neighbors.

It is also possible that you have a malfunctioning Z-Wave device. For example, the Aeon Labs energy meters are terrible Z-Wave neighbors, they can be very chatty on a network and if paired securely are almost guaranteed to kill Z-Wave. If you do have one, remove it and see if it improves the rest of the network.

They're the iris ones. They have the z-wave repeater in them and they're all paired afaik. They show up in hubitat as "Generic Z-Wave Repeater". I have 5 of them out of perhaps 165 devices though at least half of those are zigbee.

Ah I found your device handler for these. It looks very helpful, thanks!

1 Like

Interesting 3/5 of them pass the communication test. Of the other 2, one is out in the garage where I've had a few problems, and the other is... misnamed at this point. I'll have to find it.

OK intersting. So both of the ones that aren't communicating are sort of at the edges of my installation. That said, there are wired GE z-wave outlets within 10 ft of each. I would have thought these would have made good neighbors in extending the mesh to these repeaters but that doesn't seem to be the case.

I'd also like to draw your attention to something that helped me (when I had zwave issues).
I found that just rebooting didn't really do anything.
But, rebooting and waiting 5-10 minutes before turning on the system did clear up many issues.
Also, if your database has been corrupted, a "soft reset" is in order.

In the "old days", when I wanted to re-pair my locks (for example), I would just pull the power cord from my hub, and bring it close to the lock.
NO GOOD!
Abruptly pulling the power can cause the database to become corrupted!
Use the Shutdown command!

1 Like

@jtmpush18 thanks for the advice. softreset = shutdown? a shutdown followed by a 10m outage then? I'm certainly willing to do that. But not at night :wink: Tomorrow morning.

No soft reset is basically a restore of a fresh database copy. This will fix any database corruption issues. You do this via IP:8081 port. Make sure to get your hubs MAC as you will need it.

I had a similar zwave issue and it ended up being a bad device. But finding it was a process and @bobbyD was able to look at logs to help identify it.

Alright, well I got my 2 non-working repeaters back online after reinstalling.

So the soft reset procedure would be:

  1. Take a backup
  2. soft reset
  3. unplug for 10m
  4. reboot
  5. restore backup?

I won't lose anything that way?

After you soft reset and it comes back up to the getting started screen go to the bottom and it says restore from backup. Do that and load your backup. Then it'll go back to getting started. Ignore that and go back to port 8081 and perform a reboot. Then the hub will come up as normal. I feel like you should let it come all the way up before you do a shut down for 10 minutes. IMHO

OK well the magic incantations are completed. I guess I'll see...

I backed up, restored, a slight change here- it doesn't automatically go to the getting started page, it just says the restore was complete and stays on the restore page. I then went to 8081 and rebooted, waited for it to come up, then shutdown for 10m and restarted once again.

All of my repeaters are online, including the 2 that were having trouble. Locks seem to be working with reliable lock, though unlocking with alexa isn't responding. I'm guessing that's a different issue.

UPDATE: Everything seems to be working properly, no slowness out in the garage which was an issue. MQTT integration to Home Assistant working without issue. Locks are working. Web interface to hubitat and dashboards seem to be loading faster. All repeaters are online.

Thanks for everyone's help. I never would have hit on this solution on my own.

1 Like

See @bill not so bad. Glad you didn't lose faith.

Glad to hear that everything is working again!

I think that even "tech savy" individuals can unknowingly corrupt their database.
I think that the more we tinker, and optimize, and try different devices, the more likely it is to happen.
I think that we can get clues that it's happened it by looking at the logs.