One of my C-8 Hubs has become inaccessible on multiple occasions.
This is one of 4 hubs in my network, 2 ar C-8 and 2 are C-7. This is the only hub that is connectd via WiFi. The others are connected via Ethernet. This hub is in a building with no wired Ethernet.
The Hub are NOT in a Hubitat Mesh, they are integrated with Home Assistant using the Maker API.
This particular hub has gone inaccessible on the primary web UI (and Maker API) interface 4 times in the past two months. I have rebuilt the database once. I started rebooting the hub nightly to see if that would help. The problem persists. I don't have this problem on any of the other hubs.
Platform version: 2.3.8.122
Hardware version: C-8
How is it connected to the LAN?
Ethernet / Wifi : WiFi
Static IP set on Hub or DHCP: IP set via DHCP
DHCP Reservation set on Router/Server? Static IP set via reservation on router
How is Hub being powered? Supplied power block
Brief explanation of network topology.
Mesh Wifi Network (Ubiquiti)
Multiple Buildings interconnected using Ethernet and Wifi
Single Broadcast Domain (VLAN)
Brief explanation of firewall / DNS / DHCP services.
OpenWRT firewall provides DHCP and DNS relay services
Is Jumbo Frames enabled anywhere on the LAN?
- No Jumbo frames are disabled on all WAPs and Switches.
How many devices total (approx) and mix between Zwave, Zigbee, LAN/Wi-fi
- On this hub: 6 Z-Wave devices
What user / custom integrations are you using for LAN/IP/Cloud based connections?
- Maker API (integrated with Home Assistant)
- No other integrations local to the hub
Screenshots of Past Logs from the incident, and also of any errors being generated recently.
- The logs show the device stopped working at 8:06:46 last night: app:22024-03-03 12:06:51.493 PMdebugdevice event: {"name":"systemStart","value":"2.3.8.122","displayName":null,"deviceId":null,"descriptionText":"System startup with build: 2.3.8.122","unit":null,"type":null,"data":null} app:22024-03-02 08:06:46.258 PMdebugdevice event: {"name":"voltage","value":"121.408","displayName":"Guest House Heater","deviceId":"2","descriptionText":"Guest House Heater voltage is 121.408 V","unit":"V","type":null,"data":null}
can you put a kasa wifi switch on the hub.. then at least if it is unreacheable if the switch is up you can reboot it and then do a reboot with restore database to get rid of any corruption from the forced reboot.. i keep these switches on all my hubs because some are in remote location..
I should point out that the Diagnositc Tool interface on port 8081 stays up when the hub goes otherwise unresponsive.
Getting Ethernet out there will be challenging, but I may be able to swap the 2 C-8's to see if the issue follows the hub. The lockups are unpredictable so I will just have to wait and see.
The Wifi switch is a good idea, luckily the Diagnosic interface has provided a way to remotely reboot so far.
I have a C-8 hub which has frozen 5 times over the last month or so. I upgraded to the latest software version last Saturday after it hung. I was hopeful since the release notes mentioned a possible fix for hanging. No go. It hung again today. The hub is hardwired to Ethernet.
Fortunately a power cycle of the hub gets things running again. I have been using Hubitat Hubs for many years and have never seen this kind of problem before. So please, software guys, fix this problem.
I just realized the amount of work involved in swapping the 2 C-8's. The other C-8 has tons of Z-Wave devices that I would need to reinclude. Some are behind heavy appliances. So, I'm afraid I don't have a good way to troubleshoot this with a wired Ethernet connection.
I have a drawer full of C-7s, I'm wondering if I should just swap it out for one of them with a Wireless dongle until the C-8 platform is more stable?
Do you have the Sonos Advanced software and drivers installed? In tweaking his app, the author has made several changes that have taken my backup hub down many times. Each time, the diagnostic interface has always been available and I was able to reboot... Just throwing this out in case you might be using that app...
EDIT: I also have WiFi based Govee plugs on all my hubs, remote and otherwise so if I "have" to pull power, its easier than actually pulling power...
I would always go with the reboot though if you have access to the diagnostic tool. Pulling power can cause other issues.
The Powerline adapters are an interesting thought, though a bit pricey to use for this. The Wifi coverage in that building is great, and I got the C-8 specifically since it has Wifi. But I have stuck this in the back of my head if it comes to that.
No Sonos drivers, their are no special drivers on this hub, the only app is Maker API. And it just has 5 Z-Wave devices (one with S2 security the otehrs with security turned off). So, it's really non-taxed and under utilized.
It's main purpose is to manage the Kwikset Z-Wave lock in the building -- which was problematic through the Z Wave mesh but works flawlessly if a hub is within 5 feet of the lock. So, when the Hub crashes the automation around lock codes goes with it, luckily I have been able to reboot from the diag tool so far -- though Wifi plugs are also sitting in the back of my brain as a potential need (unfortunately).
Stop rebooting every night, that is probably doing more harm than good. If you find the hub is getting low on memory or something that would need a reboot I would schedule it at most once a week. As you noted it usually does not prevent from platform from crashing (unless it is specifically memory related).
6 Z devices + Maker should definitely not crash the hub platform.
Does your mesh network have any sort of "Channel Optimization" enabled? This would be where it scans all the wifi channels and then adjusts the broadcast channel to use the one with the least noise on it. That has been known to knock the hub out, but I would expect the diag tool to not be accessible in tat case so it seems unlikely.
Maker API is only being used to interact with the devices on that hub correct? You are not pushing other devices to the hub from HA (which I dont think you can do with Maker but just confirming)?
You could try turning on debug logging on Maker API just to confirm it is not having some sort of obscene amount of data being run through it.
D you have hub protect with cloud backups enabled? That has been known to crash Z-wave, but usually not the hub platform.
Other than that there is so little going on with your hub, there is not much to recommend. As a last resort you could PM @support_team and ask them to review your engineering logs for more clues.
I only started the rebooting every night when this problem started happening. I have stopped doing that since it didn't help. Now the other C-8 has crashed twice, again the admin interface was still up so I could reboot it. This hub is connected to Ethernet so no Wifi is involved.
The 2 C-7's I have have never had this issue, so it seems to be something on the newer platform.
The 2nd C-8 is a little bit busier than the other hub. It has 26 Z-Wave devices and Maker API is the only app.
I think I may switch these to C-7s and see if that stabilizes things.
