Home hub for lan devices only

So hear me out on this. I know it seems a little counter intuitive to what Hubitats as much local as possible design directions, but I think this could add value. I am really looking at this as being a device for homes with multiple hubs or simply environments that stear away from zwave and zigbee and primarily stick to loval lan only.

First the reason I have even thought about it is all of the talk for folks that use second hubs and why. It seems there are good reasons for it. Even as i continues to expand my setup i am seeing it more and more. For me it would really be about getting seperation between what appear to be power hungry cloud or local lan connected devices and the lean and fast zigbee/zwave environment. The most intensive apps on my hub are by far anything to do with lan or cloud devices.

With hub mesh it should be fine to seperate them out and allot of folks already do this.

Clearly known benefits of a second hub are not in question, but would it make sense to have a version of the hub that didn't have zigbee or zwave radios. Would those radio removals reduce cost and overhead enough to have it make any sense to even provide that option. My guess is probably no but i was still curious about the possibility.

There are several community members who dedicate a hub for the uses you indicate and turn off the radios. Others use an RPi or equivalent for similar reasons, or to integrate devices not supported directly by HE.

3 Likes

Well yea.. I am aware of that. I have a Node red instance running on my home server as well. The fact that other members of the community do this with a second full blown hub was kind of my point though.

I was wondering if there is a cost effective way to for Hubitat to make a secondary product that would provide that extension and use the good integration they have with Hub Mesh and just not have extra stuff not needed to reduce cost. I would imagine there is a bit of extra electronics and equipment not needed in a C7 hub that simply does calls across the lan. That then should be cheaper to provide I would expect. Heck if the Hubitat firmware/os/Software was flashable to a Raspberry Pi you could buy a the pi and case and do it for less then $50. As I understand it though hubitat doesn't do that and honestly i don't blame them.

I don't think it would be worth the extra SKU and all the overhead it implies. The audience is so small (at least I hope us multi hub users are a small percentage of the total !!)

The other (larger) element is to repurpose. Today that Z-less hub is good for Internet facing but tomorrow you might decide to add 2 Zwave devices that are in a tricky location. 2 devices wouldn't be a big impact on the internet facing tasks. But I'm just inventing a story to trigger ideas because this Home Automation task is shifting all the time. :slight_smile:

3 Likes

You have to consider that (at least speaking for myself) as a user of multi hubs with one having no "Z" radios turned on works great in my current setup, however I also view it as an insurance policy if ever in a tight pinch that one of my other hubs fail that I can easily Turn on the Z radio and be back up and running the same day if need be. I personally would have zero interest in a hub with no radio connection ability.

1 Like

It's a $150 device. They manufacture them in bulk batches somewhere (probably China - I never bothered to look). The cost of creating a model with no radios is probably greater than the amount they would save by not putting in a radio, while requiring more time, effort, and storage costs for an audience of you and maybe a handful of others. I have three hubs and would never buy the version you are suggesting. Spend the $150 and just turn the radios off.

1 Like

I like the idea. Instead of a small box without radios, II see it as software and a license code. Docker container perhaps.
I’d move all of my IP-only devices and apps, dashboards, etc. to a RasPi or run the container on a NAS.

FWIW, in similar previous threads, Hubitat staff have given the impression that they have zero interest in this.

4 Likes

Well I am just asking about the practicality of it. I don't know what the cost of the parts that would be removed would be which is why i asked the question. I know there is a breaking point where it makes no sense. Heck even if it was just a repurposed ARM box flashed with there software it would fill what i am asking.

I am not surprised by this depending on what the context of the system that would run their software. I mean if most folks would run it on a wintel box for docker that could create allot of support issues. This is largely why i asked about using similar ARM based hardware.

I would jump on it if i had the option to load the software myself and pay a license fee for similar hardware. I have a few 4GB raspberry pi's i could load up right now. Heck even if it didn't come with anything support for anything outside of then maybe making hub mesh work, i would probably do it.

I know I can load HA on my server, and i have Node-Red already, but in many ways i would like to keep as much as i can in Hubitat software. The guys here have really done a great job with their set of software/hardware.

I think they see it as a support nightmare no matter what other hardware the platform could theoretically run on.

This gets to the practicality issue. There will probably never be enough people in their target market(s) for them to spend any serious time or effort on something like this.

3 Likes

That is why I initially asked about a hubitat hub that is a lan only. If they have control over the hardware the support risk over their current one should be minimal. Hub Mesh needs to be really solid for this to work though and my understanding is that it has come a long way.

Practicality of it isn't lost on me. I don't know how many folks have enough devices that would need it.

What also isn't lost on me though is also the large impact of LAN/WIFI/Cloud based devices. My biggest cpu consumers are Roku TV's, Ecobee thermostat with Ecobeee Suite, Google Chromecast devices, Wemo Connect and for 2 devices, AWS Wether device driver. I have also setup Hub info and the Hub Ping drivers.

The hub ping driver seems to consume a good amount of resources when it is pinging. If I setup a few different monitors with a fairly small interval for checks the cpu on the hub would skyrocket. The roku TV app depending on what it is doing uses up to 3% cpu per roku TV device in use. The Ecobee Suite with the Ecobee thermostat which is 100% cloud based is probably one of the biggest single consumer with it's helper apps. This usage can be over 3-4% of overall usage for just that one app. It isn't hard to see how that can add up quickly and someone can start to see major impacts on reliability.Most of what we see is 5 min interval usage. Interval usage though great doesn't help really understand the usage accounted for from spikes that occur. Functions will cause fast spikes in usage that that can cause unpredictable behaviour or slow actions.

This is also where I think this kind of device(or well using a second hub to offload any lan/cloud) could help make your environment more reliable. Well atleast the zigbee/Zwave side that is. So support for these devices could actually be simplified.

Understood; the demand for this configuration makes it unlikely to happen. And @waynespringer79 pointed out earlier, there are advantages to getting an extra C-7 and just turning the radios off.

1 Like

But doesn't the hub Protect subscription somewhat negate the benefit of a backup full blown hub.

I also don't really agree with the idea that someone should simply accept buying more then they need for the task. It wouldn't be hard to create C-7 Lite hub. This concept is done all over the place. Now if we are such a small population that is a whole different situation. I also don't have any clue as to how big the customer base for Hubitat actually is either so there is obviously that as well.

The original C3/C4 hubs were similar to what you wish for.
I have a number of them that I only use for lan/cloud/processing
They originally used USB sticks for the z-radios but they work well without the sticks plugged in.

I run my C-3 hub w/o the Z-Wave radio plugged in. It works great, except when it must be rebooted. It will not reboot w/o the radios. Any time I apply a new hubitat update, I have to unplug the hub and plug it back in after waiting a safe 5 minutes for the update to apply.

If I plug the radio in, it doesn't have this problem. So, since the problem is more prominent now with the new versions, I just plugged my radio back in, and just don't use it.

Having a hub for LAN integrations only and using hub-mesh to access those devices from the main hub has worked out really well. The hub mesh is near instant. And when LAN devices go offline, and the hub is busy dealing with a timeout condition, my automation hub just hums along w/o issue.

I am now considering the potential for one of those cheap TV streaming appliances from china that ship with Kali Linux to re-purpose as an inexpensive integration host for things that run in node.js, like home bridge, so as to not use my NAS for that task....

Still, if Hubitat ever released something like this, I would buy it on release day. Until then, I just wait for sale prices and buy a new hub. :slight_smile: I now have a C7 for my motorhome, and C7 and C3 for my home....

1 Like

Did you turn off the zigbee and z-wave radios?

BTW... here is my motorhome...

2 Likes

@Cobra

What are you using to monitor all those data points? I want to do this exact thing. And I want to be able to generate alerts when my holding tanks reach ¾ full, so I was playing with the idea of installing some custom monitoring into an Arduino, or Raspberry Pi, but I just learned there are BT enabled control / monitor panels... Low power is the key with this though...

I do this in a number of ways

I have a C5 hub in the MoHo
I have 'Shelly Uni' devices monitoring the voltage
ST leak sensors etc..

These are connected to the C5 hub

I then wrote a couple of apps to collect the data
I can send the data back to a hub at home via makerAPI (the 1st app)
MQTT is what I'm using in the new app - this is much better but you need an online broker
(I'm lucky I have some public IP addresses so mine is at home)

This is using MQTT back to my node-red (on a Pi4)
The dashboard is from Node-Red

There are a couple of things to note
You need an 'always on' internet connection in the MoHo (I use a cheap £20 monthly unlimited data sim from 'Smarty' but you can get limited stuff from them for £6 - But I stream into my MoHo using the same sim - Obviously I'm in the UK)

There are some overheads in terms of power used.. I haven't really measured for a while it but I would expect it to be only a few hundred milliamps
I use a couple of solar panels to keep my batteries topped up and don't really have any problems

Of course if your MoHo lives on your drive, or you just want the data when in the MoHo,you won't need all the internet stuff

2 Likes

BTW
You'll notice from the picture that the mode is 'Home' and that the voltages are quite high
This is because the MoHo is currently on the drive (Normally in storage a few miles away) and connected to the house power
Mode goes to 'Occupied', 'Unoccupied' or 'Storage' when no at home which affects things like my internal camera and motion sensing etc
For example when occupied the camera is off :slight_smile:

1 Like

I have a C4 hub, without the USB Zigbee/ZxWave radio stick plugged in, that I reboot for Hubitat firmware updates without any issues whatsoever. I did make sure to disable both radios in each one’s respective settings page.

I have been running this C4 in this configuration for years. The only difference between the C3 and the C4 is the silkscreened logo on the case.

2 Likes

Download the Hubitat app