I didn't experience this on the upgrade myself. Though early on the only way to get around a corrupt pairing was to delete the home in HK without adding a new home. Remove the hubitat integration from HE, reboot the hub, re add the integration then change the QR code then re add the bridge. Dunno if that will help or not. (I also assume your HK and HE are on same subnet)
What is the home hub used? It would be under Home Settings - Home Hubs & Bridges. The current one will say Connected.
In my limited experience, Apple TVs work better than HomePods. Sometimes a HomePod restart is required before it starts talking to the hub again. Also, are all iDevices and home hubs (Apple TVs/HomePods) on the same OS version, like 16.2?
@gopher.ny might have a solution. There seems to be a known bug that i encountered unknowing. Will look into that and see what happens after that change.
2.3.4.130 got the fix. I mean, there could still be something else, but the particular issue with integration tripping on non-numeric characters in the power attributes on the outlet devices should be resolved.
Ugh. I’m down again after update to Platform Version 2.3.4.130. It was good for about a week. Restarted integration, rebooted hub, relaxed security is on. I don’t know if I can handle another full install. It annoys me that it’s a black box. There are no error logs and no way to tell what is going on. It just quits.
I’m down for the count again. Reset everything. Getting the accessory is out of compliance response from HomeKit when trying to add back HE. When I have another 2 hours to spare again, I’ll cycle through all my devices 1-by-1 to try to figure out which is the offending one this time.
I am up and running again with .130. Thx @gopher.ny
Is there a short reference guide somewhere on recovering from these types of issues? @rlithgow1 mentioned a few steps to take, presumably from experience?
I am aware that this is all beta, is there a structured way to provide feedback on beta’s or a beta programme? Currently im winging it as i encounter issues, possibly flooding the community with issues that might be captured somewhere else without me being aware…?
@brad, will excluding it from the exposure list to HomeKit from the HK integration app not be sufficient?
The best way to get someone's attention during this beta is to form the subject of the thread as [2.3.4.xxx] C7 (or C5 depending on what you have) Insert problem issue here.
List it under integrations board. This is an open beta vs the closed beta section (which you can join, just send an msg to @bobbyD to participate) Though for the purposes of this open beta, it's best that everyone sees it and can chime in and give more datapoints if possible.
Yes, this worked for me last time… and again this time. For some reason, one of my contact sensors started causing the issue.
@mike.maxwell - This issue was caused by my Third Reality contact sensors on Generic Zigbee Contact Sensor (no temp) device type. It started throwing a 200 for battery level (we have discussed this before), which I’m assuming is what crapped out the HomeKit integration. I have moved all of my Third Reality motion and contact sensors to the Generic Zigbee drivers that are not the “(no temp)” type, and it brought my HomeKit integration back online. These drivers seem to be quite an issue. I’ll monitor using the other drivers and see what happens.
@gopher.ny - Is there a way to keep incorrect values from crashing the integration, or at least providing a report on what is causing the issue? I have Hubitat Homebridge v2 running on Homebridge in my other house and I have NEVER had these kinds of issues. There, I see errors for invalid values (like thermostat set temp too low) in the Homebridge logs, but it seems to filter them out so they don’t crash HomeKit.
No email required. @H_Local if you'd like to provide more detailed feedback, I recommend joining our Beta program. To do so, please check out the following post:
@brad the strange thing was that even the bridge instance in my home assistande configuration had issues. That was resolved too after this, so it seems a structural issue in the current beta.
I second your suggestion to create a filter for these issues and capture or wrap them.
The implementation has a generic way to deal with values that are out of range, and reported battery level of 200 shouldn't be an issue. Can you PM me the hub id it's happening on, I'll check the engineering logs?