No, but would make sense that they are. I only have my main InfluxDB and Grafana instance hooked up to Maker API / Node JS linked to my HE hubs.
Unless you’ve got an SSD drive connected to your RPi I wouldn’t recommend running these two applications to process data for a dashboard. That’s a lot of read/writes for a microSD card and they typically won’t last long.
1 Like
Great stuff, thank you everyone; I have now enabled the required entities in Home Assistant (HA) 
.
Note that I am running HA on a RPi (love the addons such as Adguard and ESP32/Zigbee2MQTT), and use it to bring a few devices into HE via Zigbee2MQTT and HADB. And, similary to bring some HE devices into HA via HACs just to view logs and graphs only (not for automation) which are supported natively by HA.
I have tried various graphing solutions, including Hubigraphs and webcore and fuelstream (influxdb & grafana just seem too complex), and just decided now that the native graphs in HA are enough for my needs as it requires no setup and tinkering.
I have Grafana setup, but I really don't use it that much. If I want a quick glance at what's going on I use an apex-chart card.
2 Likes
Have you tried Quick Chart? A custom app on HE.
Simple but works locally on your hub
I thought Quick Chart did not run locally. From the first post in it's thread:
1 Like
My bad. What I meant to say was that it requires no other hub or computer. Just install the app on HE.
1 Like
Over the course of a few years of working with Hubitat and many CPU crashes and Zwave radio crashes, my frustrations had lead me to read about @thebearmay 's Hub Info device/driver months ago. I don't know why these hubs are so prone to crashing - my desktop computer never crashes, even if its CPU is completely maxed out.
I've got a lot of experience writing rules in RM, but some of the concepts in these threads about Hub Info, like endpoints, stacks, queues, http posts and gets, are like a foreign language to me.
I apologize if this has been answered before, but can Hub Info V3 reboot one hub that is part of a Hub Mesh, from a second hub that is part of the mesh, if the instance of Hub Info on the first hub is a shared device on the second hub? If so, can Hub Info reboot the first hub from the second hub if the first hub has completely crashed and is no longer addressable in a web browser at its IP address?
No... As the hub itself isn't shared. Just write the rule on the hub in question.
Now that said, you should be investigating why the hub is "crashing" (Note, pc's crash all the time, usually because of a software or hardware issue. The key is tracking it down).
When it crashes can you get to port 8081?
Do you have Jumbo frames enabled anywhere on your network?
Rogue device(s)
Power metering can crush a mesh and lock the hub...
etc...etc... The key is to work through tracking down the source.
Hub has to be addressable via HubMesh, and the Reboot Option enabled on the remote hub, but the reboot command could be initiated via the mesh.
3 Likes
My C-5 never crashed, my C-7 never crashed (well, except for the one time my wife unplugged it while I had it on an extension cord to pair a door lock, and she wanted to vacuum), and my C-8 has never crashed. They all have been, and are, on a UPS with a shutdown rule triggered by Ring Extenders. I understand your frustration if yours are crashing. If your hub is on a UPS, then perhaps your hub has a hardware issue. The only issue of which I am aware that crashes the hub is if you have jumbo frames on your network. You ought to investigate.
But, no, these hubs are not prone to crashing, from my experience over the past four years, ever since I migrated from Wink (with a brief two-week interlude on SmartThings after Wink).
3 Likes
Hmm, I guess you could do a post command too
Updated to version .146 about 45 minutes ago and the hubAlerts : [platformUpdateAvailable] will not clear.
I've tried refreshing, configure, initialize and rebooted twice.
Hub Monitor continues to raise an alert about a platform update.
Any way to clear that?
Set Zwave Status & Hub Alerts to zero and turn Remove unused attributes on and hit save preferences. Click Refresh - at this point the hubAlerts attribute should no longer show, and you should be able to set the preferences back.
@thebearmay thank you for your answer to my rebooting via HubMesh question. I sometimes have situations where one hub has crashed and find.hubitat.com won't connect, but the Hubitat diagnostic tool will still connect and enable a reboot. In this case, the crashed hub is not addressable by the second hub on HubMesh, so I suspect based on your response that Hub Info won't reboot the crashed hub. Do you know, if I am only able to access the crashed hub via the Hubitat diagnostic tool, is there any way to initiate a reboot on that hub from a rule on the second hub?
Also, you stated that the preferred method to install Hub Info is with HPM, but I'm so used to installing drivers by copying and pasting the code from github into the Driver Code/New Driver field and clicking Save that I did it that way without much hesitation. I created Hub Info devices on my hubs which appear to be OK, but I didn't Initialize them yet. Am I asking for trouble, and should I remove the driver/device and reinstall using HPM? Thanks.
If port 8081 is responding then HubMesh may be responding also - no guarantee but maybe. As to the install method, if you’re comfortable with the manual install there is no reason not to do it that way.
@rlithgow1 thank you for your answer and advice. Yes, I want to try to track down what is happening. I don't care about monitoring power usage, so in general I try to disable power reporting in my devices. Is this what you were referring to?
For simplicity (although maybe not ideal) I use my wireless router as a network switch, as it has multiple LAN ports and will control LAN-side network traffic. I'm sure a separate network switch would be better. There are no settings for Jumbo Frames or LAN packet size on the router UI (although there is a setting for WAN transfer rate). Searching online, it appears to have a default setting of 1500 for MTU so I doubt Jumbo Frames is being used.
I should mention that I get a hub crash about once every three weeks. @thebearmay stated in a different thread that the memory usage on Hubitat hubs steadily increases over time, and I don't regularly reboot these hubs. I'm starting to think that maybe this is a cause of many of my crashes, and I see that @thebearmay recommends a reboot when a hub gets below 90 to 100 KB of available memory. As you can see, I am new to Hub Info.
That doesn't explain the Zwave radio crashes though. Compared to most people, I am sure there is far more radio traffic on my hubs. I use Zwave automation to monitor and control environments in different rooms for my reptile breeding profession. I use multiple temperature/humidity sensors for example, one of which I have set to report every minute in a room that needs very tight temperature control. I'm probably seeing one to two Zwave transmissions per minute on average, so there must be times statistically-speaking when numerous devices transmit at more or less the same time. Based on what you were saying about power metering, I'm thinking this radio traffic could cause a Zwave crash.
Also, to answer your other question, the vast majority of the time when I get a crash, I can still access the crashed hub on port 8081. I'm wondering if there might be a possibility to create an HTTP POST as an action in a rule on the second operational hub on the Hub Mesh that would run the Diagnostic Tool and initiate a reboot on the crashed hub, but I haven't tried this.
1 Like
@672southmain, thank you for your input. I've been planning to put both hubs and a network switch on a UPS, but I haven't done that yet. Not only would that save some work/grief following a power outage (which is fortunately rare where I live) but I suspect that a UPS would provide some line conditioning and therefore be safer for the hubs and their mini-network.
1 Like
Sorry @thebearmay for contaminating this thread further. This is my reboot rule (3 hubs). It is activated by another rule turning on a virtual switch that automatically turns back off. I also share the switch to Homekit for initiating the reboot via Siri.
Good plan. It’s possible that your database has become corrupted by the power fails, and that could explain the crashing. The way to fix a corrupted database is to backup and download your database, do a soft reset (which wipes the database), then restore the backed-up database. That process purges corrupted records.
