Things are pretty fast these days. I just broke the news to my wife about what's about to happen. This'll be fun - let's see if I can do it without anyone noticing.
2 Likes
I am usually one to ask for forgiveness afterwards.... I guess there is no right or wrong here either.......
4 Likes
When it comes to HubConnect topologies there are no right or wrong answers. Use what works for you.
This is exactly how I started off... One hub in the basement, one hub on the 2nd floor. Dual radio stacks on each (Zigbee/ZWave) with all devices shared to the server hub. I've since added a 3rd hub to the mix for critical devices and LAN integrations. The theory being if I keep the hub to under 50 devices it will hopefully not suffer any showdowns. (50 devices is just an arbitrary limit)
You won't have to worry about latency for a long time. I've got 555 devices on the server with about 500 coming from remote hubs. Even for Zigbee and Z-Wave devices I see response times consistently under a second, which is acceptable. I do see some extreme latency in 2-3am hour which I am told is due to scheduled background maintenance being done on all of the hubs, but outside of that it's always instant.
As devices grow into the hundreds, using websockets in both direction can become detrimental to the remote hubs since they receive and have to parse all events raised on the server. HubConnect 2.0 is going to have a solution for this which will be benficial to larger systems.
You are a brave man.
I have to carefully assess risk first. Usually I hope she doesn't notice.. For example, over the weekend I mounted the 2 motion sensors in her hobby room. A few years ago we agreed to leave them on a shelf because they looked ugly. FInally got tired of seeing them knocked over or facing the wrong way so they're both stuck to the wall not.
She has not noticed... Yet.
2 Likes
I hope you have your "excuse" ready for when she notices. You know if is a matter of "when" and not "if". Based on my experience, something "useless" being knocked over is not a valid excuse.......
Good luck and may the schwartz be with you!
1 Like
After 19 years there's not much that gets by her... She's probably already noticed. It's just a matter of when, or if she even brings it up. And that is decision is likely dependent on what else I do wrong in the near future.
She's definately a Druish princess.
4 Likes
Well I'm taking it slow y'all. Before I tear apart my meshes I'm setting up my 2nd hub as the controller first. So far I've got Google Home, Chromecast, HSM, and Mode Manager running on it. Next I'm moving the Dashboard over and I think that'll do it for the heaviest apps for me. I might put the LIFX app on the controller since it's 'different', but haven't decided yet. It looks like it doesn't really matter where I put Lutron, but I'll probably just keep one instance on the controller.
@srwhite Amazing work here - instructions were clear as day, install was flawless, and things are working as expected. Thank you!
My plan is to begin moving over just the global admin-type rules next. I do have a third hub (with stick) that I was planning to keep in a drawer as a "spare"... But... We'll see what happens. 
Feel free to install the Lutron app on all your interconnected hubs. It makes the button integration much simpler.
1 Like
ok, i'm back. i just spent the last 24 hours trying to recover my hub. it seems to be recovered, but this problem is unchanged. same errors.
@dan.t the "clientURl" is http://192.168.20.33:20009/ which is the correct IP address. again, the info seems to be updating, it is just throwing tons of errors.
@srwhite I can't generate a new connection key as I can't actually open the server instance. doing so just brings me to the page that says click done to finish install, but that doesn't actually work. I did create a new server instance and put that key in my homebridge config and all that did was delete all my room associations with my devices, I was still getting the error.
Can I delete the duplicates then? I assume so...
I started having the same problem. Everything works perfect, Home is able to control everything and updates immediately, but I can’t open the Homebridge server instance to add or remove devices. I only have the option of hitting done as it thinks the app has been updated. No problems opening the other HE instance or the ST instance.
I have my Z-Devices split by Floor.. upstairs and downstairs. I split my Pico's too. All hubs have the Integration, the buttons are available to any/all hubs but due to lazyness, I only tell each integration about the picos that might be used. Technically I could define a Pico on every hub and have one button press do something different on every hub.
Well that makes sense. No sense in cluttering things up with things you don't use. I am the same way. Plus, I can see this get confusing real fast if each hub isn't easily distinguishable from one another. Maybe one day we'll have color schemes. 
Check the logs.. Something is failing to properly initialize.
I'm getting this error in homebridge also
(node:44) UnhandledPromiseRejectionWarning: #<IncomingMessage>
(node:44) UnhandledPromiseRejectionWarning: Unhandled promise rejection. This error originated either by throwing inside of an async function without a catch block, or by rejecting a promise which was not handled with .catch(). (rejection id: 2)I need the full log, just those single line don’t tell me enough. Are you on the latest version of the Homebridge plugin?
here are the last 500 lines
edit: deleted logs
and yes, most recent homebridge plugin. I just recreated all my docker instances.
Do me a favor, open a browser and go to
And click the “Download Log file” button. I’ll pm you an email address where you can send the log file to.
I got your log file. Thanks for that.
I have an assumption... when you go in HE into the HubConnect Homebridge instance, what is the URL that is displayed in the browser. I am interested in the app instance number and want to see if it is 483
yes, it is 483
I am seeing a ton of "Server Error 500" message. You see updates of the devices on iOS because the plugin connects to the Hubitat websocket and retrieves the updates from there. But it has major problems communicating with HubConnect.
@srwhite, @csteele any ideas here? I can see that the plugin calls everything as usual. Every http request to the HubConnect instance ends with a HTTP 500 error...
@jrau272, you don't see any other issues with the hub correct? I am wondering if you have a problem with the DB on the hub.... We have seen these Server Error 500 before but they would appear more widespread and not contained to a single app. Unless that single app has an issue... In the latest version of HubConnect, @srwhite added a technical report. You can get there by going to
HubConnect Server -> HubConnect Utilities -> Technical Support Report
It might be beneficial for @srwhite, @csteele or me to see that. If you could get that one to us (pm, don't post it here publicly!) we might be able to see more. Something is clearly not straight
