Habitat C7 hub keeps going offline

I have a habitat C7 hub that I use for managing z-wave devices and exposing them to Alexa and Google Home VAs, but most of the automations (and Zigbee/Matter devices) managed in HomeAssistant. This was working well for a few months now, but for the last few weeks, the Habitat Hub keeps going offline and I am unable to access any of the z-wave devices either via Voice Assistants or from HomeAssistant. Only way to recover is to do a network reset (using the reset pin on bottom) or a power off/on cycle, but it goes offline again at what seems like random intervals.

At first I thought it could be due to daily 2am backups and disabled that, and even used the diagnostic tool to do a soft reset and restored from backup. Everything works fine until hub goes offline. Logs do not show anything unusal at all before it goes offline, and they start writing once I reset.

Any help in figuring out how to make this more stable is greatly appreciated.

Does diagnostic tool work when it is offline?

:point_down:

1 Like

When hub is offline, I am unable to reach it at all and can't launch diagnostic tool either.
The network seems to be down, and my router (unifi) shows that the device is offline.

Please provide responses to everything else that you can in my post above.
It will prevent a bunch of back and forth questions if you just answer it all up front.

3 Likes

I have performed these steps already and it is still having same issue.
Model of Hub: Habitat C-7
Hub Platform Version: 2.4.1.177
Z-Wave version info:
VersionReport(zWaveLibraryType:7, zWaveProtocolVersion:7, zWaveProtocolSubVersion:14, firmware0Version:7, firmware0SubVersion:14, hardwareVersion:1, firmwareTargets:1, targetVersions:[[target:1, version:7, subVersion:18]])

Hub is connected via ethernet cable, and has static IP assigned in the router, but hub is set to DHCP

Hub is being powered by USB port on CyberPower UPS, which is rated for 2.4A at 5V

Network topology: ATT ONT --rj45--> Unifi UDM SE --10G SFP+--> Unifi Pro Max 16 POE --RJ45--> Hubitat C-7
All other devices fork out of Pro Max switch, and I tested from hard wired PCs, Mac, and Wifi iPhone.

  • Is Jumbo Frames enabled anywhere on the LAN? -- Not sure
  • How many devices total (approx) and mix between Zwave, Zigbee, LAN/Wi-fi - about 25
  • What user / custom integrations are you using for LAN/IP/Cloud based connections? -- I have the HA connected using API key, and only have z-wave button controller and voice assistant integrations. on c-7
  • Here's a snippet of logs that shows the logs before it went unresponsive and after I network reset and rebooted. (Reboot happened around 7:50am and it went unresponsive after 6:22am)

couldn't upload any files, so here's a screenshot of logs I posted in Imgur: https://imgur.com/twU7JOC

Couple of suggestions.

You should definitely get the z-wave firmware updated. I believe that is showing you have firmware 7.14 and gateway 7.18. You want both to be 7.18 on the C7

I would test with a standalone USB-A power block if that is feasible. Also be sure the PoE is disabled on the port the hub is connected to.

Which HA integration are you using? The one that shares Hubitat devices to HA via Maker API I assume?

Are all 25 device Z-wave?

What is app 97 in logs that is making a lot of debug logs?

What is device 133 that looks to be connecting to the websocket?

I would test with a standalone USB-A power block if that is feasible. Also be sure the PoE is disabled on the port the hub is connected to.
Will try. its a pain to get the cable all the way to the rack as the usb cable I have barely reaches the front of ups in a shelf next to it.
The POE is automatically disabled if device doesn't ask for it, but I can explicitly disable too. will do this.

Which HA integration are you using? The one that shares Hubitat devices to HA via Maker API I assume?
Yes

Are all 25 device Z-wave?
No, some are Wifi (Shelly and Kasa). Probably about 18 are z-wave

What is app 97 in logs that is making a lot of debug logs?
What is device 133 that looks to be connecting to the websocket?
Not sure, will have to research. The hub went down again :frowning: , so will need to reset it when I get a chance this afternoon, and get to this.

You can just click the info or debug colored box on the log line to open the associated device/app page.

What Shelly devices, and what drivers are you using?
What Kasa devices and what do you have the polling set to?

Drivers or apps hammering on the TCP stack on the hub could possibly be knocking it offline after a while.

I disabled Kasa integration for now to see if helps.
App 97 is the maker api used for HA integration, and device 133 is the HA device bridge. I disabled debug logging on the maker api to see if it helps, but this setup worked well for 3 to 4 months now... just recently started seeing this error, and got really worse this week. Today it went down 3 times.

So you also are using HADB? How many ha devices are you bringing over?

After the hub has been up for a bit get screenshots of the device stats and app stats tabs from the logs page

HADB has potential to crash the hub if an extreme amount of events are coming over

I attempted to build automations in hubitat at first, but went the other route. I am not using any HA devices in Hubitat at all. I actually forgot I had that enabled. Will disable that.

But so far, I have switched to a power adapter, disabled kasa integration, disabled debug log on maker api, in HA hubitat integration, I disabled polling for status, and network reset 3 times since I posted this and the hub keeps going offline. Now it barely stays online for an hour if at all :frowning:

If you have any sort of NAS or server it is possible they are using Jumbo Frames. Do you have any old "dumb" switches around? If you put Hubitat behind one of those it will shield it from any Jumbo Frames (just as a test).

There is no need to network reset the hub to restart it. Its not a "safe" shutdown so there is no advantage from just pulling power. I would just yank the power cord out for 30 seconds then plug it back in again. That fully reboots all the hardware. You should also reboot > Advanced > Rebuild Database due to the uncontrolled shut downs.

Next things to try

  • Disable or remove HADB (app and devices)
  • Disable Maker API that sends to HA, and disable the plugin on HA (I think you can temporarily disable it without breaking everything?)
  • I also would still like you to get the z-wave firmware updated.

Thanks for the input. I will put the hub on a smart plug so I can reboot remotely. Its a pain atm to get to the rack in basement every time I need to reboot/reset :slight_smile:

I have an Unraid server plugged into the switch with 10G sfp+, so I guess it maybe using jumbo frames? I will plug a dumb switch, or see if we can disable jumbo frames on the port hub is connected to.

I will remove HADB and temporarily try disabling maker api and ha plugin if possible.

to upgrade z-wave firmware, I tried it from z-wave details screen but it failed with an error. Can you please point me to any instructions on the right way to upgrade z-wave firmware?

You are stuck in exactly the same position I was in with my C7.
It just would not fully update. I did raise it with support and had some dialog but it never got resolved.
It wasn't causing me any zwave issues but as I had a C8 I just moved my few (8) zwave devices to my C8.
This was some time ago when the update first came out.
Not much help I know but thought I would let you know that you are not the only one.

Here is what happened to my C7 when trying to update. You may be seeing the same.
Here is the last dialog I had with support when I had the problem in March 2024.

Here is where we are and what I have tried.
Shut down hub.
Power down hub.
Wait 2 mins.
Power up.
Waited 2 mins and went into Zwave page.
Hit firmware update button.
Hangs in the same place as before.
I then followed this.

1. Update hub to version 2.3.8.122.
2. Go to: yourHubIP/hub/advanced/installZipgateway715*
3. Reboot the hub from Settings
4. Go to Z-Wave Details page, then select Update Firmware to install "Z-Wave and Z-Wave Long Range 700/800 SDK 7.18.1.0"
5. Go to: yourHubIP/hub/advanced/installZipgateway718*
6. Reboot the hub from Settings

After step 3 there is no Firmware update button but /hub/zwaveVersion is accessable and shows the version that is running.
At this point I move on to step 5.
Response says it has completed and to reboot hub.

I linked the instructions above: Updating Z-Wave radio on C-7 hubs after version 2.3.8.122

thanks @jtp10181 and @bobbles
I have disabled HADB, and Kasa integrations and started with step 2 of z-wave firmware update process before i got busy with life but surprisingly the hub hasn't crashed all weekend.

I just completed the z-wave firmware update today and now its at 7.18. I re-enabled just Kasa integration. So far so good... will keep an eye and disable Kasa if it starts acting up again.

Kasa should not be an issue unless you have a large number of devices with the shortest poll time set. The author tested it at one point with very short polls and said it did work OK but was not recommended unless absolutely needed. That is why I asked about what the polling was set to above.

I would suspect it is one of the HA integrations. You could probably re-enable everything else and test, then test HA integrations last. Keep HADB disabled if not using it.