I have 3 ATVs wired and 3 HPMs. It may be a coincidence, but ever since I wired all of the ATVs, I never see a HPM acting as the Home hub. I also never see unresponsive devices outside of the occasional Hue light (on the Hue bridge) getting inadvertently turned off. I have 2 C-8 hubs with Homekit running, one of which still has Homebridge running as well with 22 devices. I would tend to think that those having issues have a router not playing nice with mdns.
Have you reserved IPs on your AppleTV or HomePods, I had problems loosing connection all the time until I noticed that I never reserved an IP on my router for the AppleTV (never needed it before using Homekit, it did not need one before like many devices). Now that I have a reserved IP, it's been solid for the last month or so.
I started having problems after updating to 2.3.9.150. Iāve tried the ping hack described in the thread. All of my HomeKit hubs and HE hubs have fixed IP addresses. No response after an hour or so and only an HE reboot remedies.
Thinking this is a HE issue.
Not seeing a lot of reports on this so either itās a 1% issue or itās local to your setup. My bet is the latter.
Well, I turned off the ping/ack app about 20 hours ago and woke up this morning with HomeKit NO RESPONSE!
First time in weeks. Turning it back on for every 6 hours.
EDIT: I didnāt have to reboot. Shortly after I manually ran the app, HomeKit started responding!
1 Like
I'm seeing multiple replies on this thread and other threads (here, here) mentioning this problem recently.
Since my last post, I have deleted my home in HomeKit, reset all HomeKit hubs, added everything back into a new hub including the ping/ack hack mentioned in this thread, ensured fixed IP addresses, ensured nothing in my UniFi network settings that would inhibit / filter internal traffic including use of VLANs, and have option in HE HomeKit app to reset hourly. Worked fine for a few days, and changed absolutely nothing about the setup otherwise. Today, I again get responses for HE linked devices and see no active connections in the admin panel.
All of this started happening after updating to 2.3.9.150. That's the only variable that changed in my HomeKit setup.
Pretty certain this is an HE issue.
Otherās situation may be different, but I think HE is dropping the link to the HomeKit hub (maybe a timeout or other reason) but the hub āthinksā it is still alive and gets no response when it tries to update the states or control HE.
I think my āpingā forces HE to āreconnectā and all is well for a while.
IF we had some logging of the connections, maybe we could prove or disprove that theory.
2 Likes
Roll back the platform version using the diagnostic tool and see if anything changes:
2.3.9.150 is too new to be the issue for me at least. This has really been an on going issue for a couple months. I can't pinpoint exactly when it first started because I don't use HomeKit that often.
If we had access to the application logs it sure would help on narrowing down that time frame and maybe pinpoint the issue. I'm hoping maybe that show a time pattern?
@thebearmay suggested that it is a 1% issue, and I like finally being in the 1%, but not in this case. If it is something common for just a few of us then we need to figure that out. I don't want to hack it by adding functions to turn off and on virtual switches. That isn't a fix in my eyes.
Here is my list of things that could be common with other people having this issue:
- Multiple Apple hubs ( 3 Apple TVs)
- Two of those hubs are on Wifi, one wired
- All devices on the same VLAN
- Unifi Network for AP and switches
- Static IPs for all devices involved
- All Apple, Hubitat, and Unifi devices are on the latest firmware
- Restarting the Homekit application does not solve the issue when it happens
- Restarting all Apple devices does not solve the issue when it happens
- Restarting the Hubitat hub solves the issue when it happens
I have to believe that this is either a device sending out of range data (battery percentage or temperature), which some have reported issues with previously, or itās the router killing the connection.
I have 3 ATVs wired and 3 HPMs.
I have 3 HE hubs, 2 with the Homekit app running. 126 devices shared on one and 31 on the other (with another 21 shared via the Homebridge V2 app)
I donāt have hourly restarts enabled.
Iām now running iOS 18 beta on iPhone and TVOS 18 on the ATVs. I have not had any issues with unresponsive devices now or in the past.
I am not using Unifi equipment. I have 2 Linksys MX5300s wired and Linksys switches around the house.
Same issue. Requires reboot.
C7 on 2.3.9.150
Shouldnāt need to send ping commands to keep connection alive.
Using Apple TV for home
DHCP for Apple TV
Assigning static doesnāt seem to repair the issue.
Logging please.
I have been trouble shooting my HomeKit / HE integration for a few days now. I have 2 homes on HE with HomeKit. Home A has dozens if not more devices and sensors connected to Hubitat C8. This home has several Apple HomeKit hubs on the network, wired and wireless, with multiple Z-Wave / ZigBee networks and several virtual devices through web API integrations.
Home B Is the exact opposite, extremely simple. 1 AppleTV serving as the Hub with Hubitat C7, both hardwired (to the same switch). There are about 10 devices on this C7 with even few connected through HomeKit. It is a vacation home so it is mainly used for safety monitoring (water leaks, temperature adjustments, door lock).
Without reasons I can't figure out, both Homes are losing the HomeKit integration daily, multiple times a day. They are all set to restart hourly and I have ran through the troubleshooting the best that I can. The Hubitat HomeKit integration doesn't appear to have any logs providing me a direction on why both of these integrations are failing.
I need to dig through this thread but I am beginning to think it is just Hubitat's HomeKit integration.
For those using Homebridge, does that require an onsite server? Specifically asking for Home B which can go a long period of time unattended without a person there to reboot, update the PC.
Yes, you'll need something like a local Raspberry Pi to run HomeBridge, if you choose to go that route.
Yes, but any small SBC or mini-PC can work, even an old laptop or desktop.
You can also setup a VPN so you can connect it to from remote if needed.
Hello!
I was having this exact issue after about a month of integrating it. I recently added a HomePod mini into my network (I was using my phone before) and it was fine for 2 weeks and the problem came up again (was showing no response every 36 hours or so.
Implementing TArmanās ping solution has kept it stable this whole week. I am using the 6 hour refresh.
Iām on a c8 on newest updates.
I know my post is not super useful to the thread in general, but this thread has helped considerably when researching/working with this issue.
Thanks!
2 Likes
Could you share a screen shot of this rule? I am terrible with the Rule Machine (I am still learning syntax and how to set it up). I have figured out the more simple rules using Basic Rules and Simple Automation Rules but the Rule Machine is a little bit harder for me to wrap my head around, especially trying to translate your rule into the Rule Maker. Thank you!
7 Likes
Nitpick really but you may want to consider making PingHome: On a delayed action (1 or 2 seconds) so that the wait sets up first and you don't risk a race condition.
I use a very similar rule and while I've not seen the wait fail, the timings have been surprisingly close.
I can not thank you enough. I have learned so much from this in the context of this Rule Machine, mainly interfacing with the tool to see how it builds the If statements. I am used to something like Excel or back in the day, C++, and this just totally helps me "see" what this tool needs to work. Thank you again!!!
