Just stopped in to say.... the title and evolution of this thread (edit: the original that spawned this comment) gives me great unease, as it might to many.
Got neither the time right now nor the tools to resolve a similar event in my environment ....and that would be consequential as I do rely on a good chunk of what I have in place for "more than just convenience" (as simple as much of it is).
Have NO idea what's wrong here, just saying "it surely is a scary and frustrating eventuality for anyone to have happen".
So I'll say this again (but plead that it not derail this thread; edit: the original one); after spending the time implementing a solution that serves a hefty number of functions in one's environment I think it's safe to say, in hindsight, that many would pay more for the "cornerstone device" that keeps it all working .
Pay more up front so that the cornerstone can afford to be robust, resilient, as self healing as possible, AND offer more diagnostic tools to help debug the kind of device and mesh problems that are "beyond it" (somewhat making up for 3rd Party device weaknesses and vulnerabilities).
Quite simply, one's time and sanity are easily worth another $100. And if new buyers don't "get that" then that's a Marketing problem, not a product pricing problem.
I realize you have some automation that is "very" important to you/yours. However without knowing why you are having issues a "cornerstone" device may not solve the problem. I will grant you additional tools can help here but different hardware may or may not be of any consequence.
think it's safe to say, in hindsight, that many would pay more for the "cornerstone device" that keeps it all working
I think "hindsight" is the one thing a new adopter doesn't have. They don't know the issues that can crop up and with reading this forum I can say a lot of us had NO issues with the current level of hardware.
I have the C7 hub which I'll assume is the lower cost hardware in your description. I have 17 Zigbee devices spread all over the house, garage and some outside. I do have a few automations that might be considered as safety (internal and external lighting triggered by a garage door opening with no vehicle in that bay). Its been working since Jan 2021 with no hiccups.
I think I get the gist of what you're saying. It's one reason I feel uneasy about the idea of using Home Assistant. As it largely leaves the user to maintain and housekeep the server itself, I'd worry that I'd mess up the installation and leave it in an unreliable state. I appreciate my HE hub for being largely self-maintaining so I only have to focus on the reliability of my own rules.
If the solution was to throw an extra $100 engineering power behind the Zigbee radio, everyone would do it, even Control4 and Savant systems that users spend tens of thousands to install.
Achieving mesh reliability is not as easy as it sounds, that's why Control4 has numerous knowledge base articles and pamphlets about "best practices" and what installers as well as users should do, or not do. Here is an excerpt from one such document that Control4 hands out to their installers, which applies 100% to every user installing a Hubitat Elevation hub:
Factors impacting ZigBee performance
Consider the material the house is made of:
Any amount of attenuation can impact the ZigBee wireless range, so you’ll need
to plan for alternate methods to get the ZigBee signal around the material.
Be conscious of concrete floors, steel reinforced floors, ceilings, walls, elevator
shafts, masonry, rock, radiant floors, cinderblock, chicken wire, reinforced
materials, such as Venetian plaster, and stucco, and so on. All of these
materials and others will deteriorate the ability for ZigBee devices to
communicate. Additional controllers and multiple meshes should be designed
into projects where these materials or situations exist.
An example of how the material can affect 2.4 GHz is shown below (source:
PDF from The City of Cumberland, Maryland which can be found here).
Check for third-party devices that are broadcasting on the same channel or all channels
causing interference:
Devices like 2.4 GHz cordless phones, wireless speakers, and baby monitors can
cause interference. An example of this can be found in KB Article 633.
Every Control4 Dealer should have a wireless scanning device to check the levels of each channel. A good device to use is Wi-Spy which can be purchased on the
Internet at MetaGeek | Equipping Wi-Fi Heroes.
When setting up a ZigBee mesh, Auto Channel is the default setting. This will have
the ZigBee antenna pick a channel at random and check how much interference
there is on that channel.
If it is clear, it will select that channel for the ZigBee mesh.
If it finds any interference, it will go to the next channel until it finds a clear
channel.
This channel check only happens once when you initially set up the mesh.
The controller does not continually check for a better channel.
There is a chance that a third-party device or devices will have intermittent
communication on the channel ZigBee is on, and if it selected that channel during
intermittent down time, ZigBee performance can suffer. Use Wi-Spy to scan for a long period of time to see this interference.
If Wi-Spy is not available, try channel 25 which is out of WiFi range.
The ZigBee-WiFi channel relationship can be found in KB Article 449.
A wireless router or wireless access point (WAP) placed on top of a controller
running the ZAP (hub) can block all wireless communication from that controller even if
the channels are far apart. To resolve this, move the wireless router or WAP away
from the ZAP (hub).
If you are using multiple Zigbee networks, it is not necessary to separate the channels by
two or more apart as previously suggested for ZAPs associated with those ZIgbee networks.
You can separate the channels by only one channel without issues. We recommend
that you use Wi-Spy or some other wireless scanning device to make sure the
channels you plan on using are clear.
Make sure the ZAPs (hubs) are not located in a poor signal location:
Find a spot that is the most central to the house or general area.
All ZAPs (hubs) should be placed near the center of the group of nodes you want to control.
The main system rack (Head-End) is generally the WORST possible location for
your ZAP (hub), and should be avoided. The main system rack location is generally in a
remote area of the home, typically surrounded by concrete walls, metal ducting,
hundreds of copper wire runs, metal pipes, and often dozens of electronic devices
producing electrical and radio-frequency noise. This is typically the last location you
want the ZAP (hub) to be placed.
If the house has multiple levels, consider creating a Zigbee network for each level to reduce
lost or slow communication.
Make sure each ZigBee device is not outside the range of another ZigBee device:
Most environments allow only a 15 to 30-foot (five (5) to nine (9) meters) range for
optimal signal strength. This can be as little as five (5) to 10 feet given the
construction materials between your devices, and can be 50 feet in open air. Know your environment well and plan ahead.
Be aware that ZigBee devices implement a robust message delivery retry
mechanism. While messages may successfully reach destinations at much longer
ranges than those described above, messages may be retried a number of times to
do so. To achieve optimal performance within the ZigBee network, attempt to
maintain optimal signal strength using the range recommendations above.
Be aware that the signal strength is only the strength it has to its first routing
node. For this reason, it is important to look for ZigBee
devices with weak signals that other ZigBee devices may be using to route back to
the mesh and onto the ZAP (hub).
As you can see reliability greatly depends on the the installer's skills to avoid interference and to know what to do when unfortunate events occur.
I guess I see some of the 3rd Party Zigbee mesh maps people post, signal strength displays, and the kind of things that you can do in analyzing a Ubiquiti network ...and naively long for just a bit more here.
I guess I am basically saying....if EVER the phrase, "we can't afford to put that level of diagnostic capability onboard HE because it would bump the price to x" comes up maybe run "x" by us an see what we think.
I claim BS on folks that feel the current price is "the top they'd pay" after having built out an reasonably sized implementation. Again, for newbies daunted by the price ...this is about Marketing. You can wrap more diverse "consequences of failure" up in an HE implementation than the home routers many folks in here paid 3 times more for.
And to the posters saying, "oh, I have a bazillion endpoints and never a problem". Come on guys, seriously, ...look under the hood of MANY consumer platforms and in the best you'll see A LOT of things built in to handle the exceptions that catch far more people than is obvious. There's a balance to be had here. Do you not remember the old days??!!!
Being able to visualize z-wave network details with the new stack in the C-7 has gone a long way in allowing end users to create strong/stable z-wave mesh networks.
It would be good if in future hub iterations, the zigbee mesh got the same love. getChildAndRouteInfo is useful, but only captures a small time-slice and is limited to end-devices and routers directly connected to Hubitat.
Same with zigbee logging. It is useful, but is only one component in the whole mix.
Many of us have xbee3s connected to our Hubitat zigbee mesh as routers and also as diagnostic tools - the amount of information they provide is similar to what the platform already makes available for z-wave meshes.
Can I strongly recommend adding an Xbee3 to your mesh. They're inexpensive (even with the current supply chain issues) and provide a ton of diagnostic information.
It all comes down to cost-benefit analysis. As highlighted above, a system using professional installers that costs thousands to implement relies on 3rd party tools to survey the site. Developing sophisticated monitoring tools, while might have some benefits, far exceed reasonable cost. In the end, reliability comes from taking in consideration the factors that can impact ZigBee performance, pure and simple.
...and here I was JUST liking the things you were saying in that post to Bobby.
Yeah, that's the prudent advice, get the Xbee.
Now I'm thinking back to the network analyzers we had to buy back in the 80 & 90s just to get some of the kinda info you can now get onboard your bloomin router (ok, yes...the network analyzers told more).
The most prudent advice is build a strong Zigbee Network that can easily withstand interference. That way you can enjoy home automation free of worries and countless hours of troubleshooting efforts.
This topic has been marked solved by the community and subsequently closed. For further discussion on this topic, feel free to create a new thread under a different category to continue the conversation.
If you are logged in, click [HERE] to get started.