Many of us are familiar with the issue of cloud backup fails and causes Z-Wave radio to go off and the "fix" is to shut down, pull power 30 secs, then power back up. We also know a fix is coming "soon" but is reliant on update from chip mfr. Got it.
I got tired of doing the shut downs, walking from one end of the house to the other to unplug then plug my hubs back in to fix this - so I "found another way". I bought two Matter smart plugs (linked to Alexa). Now I can do the shut downs (and wait for finish) then use Alexa app to shut off the plug for 30 seconds then back on. Saves the walking and I can even do it from remote if needed.
So here's my next thought - and I'm no programmer. What if we "catch" the cloud backup failure on the hub and trigger an automation that shuts down the hub, then tells Alexa to shut off the plug (for 30 seconds) then turn the plug back on?
i have kasa wifi switches they work inside hubitat via cloud or local but also through their own app through cloud..
but as a side for the 2 most important hubs i dont automatially do cloud backups.. just do one manually when i do changes or there is a new f/w release.
Hub Shutdown is a virtual hubitat device (switch) shared with Home Assistant via makerAPI and the Hubitat Integration for Home Assistant
Hubitat Power is a SONOFF Micro USB Smart WiFi Adaptor paired to Home assistant, that supplies power to hubitat. i used to have some extra checks in the home assistant automation, but i found giving the hub ~5min to shutdown cleanly before killing the power was simpler and worked better.
The 15min wait before powering back on is to allow zigbee devices to enter panic mode before powering the hub back on to help with zigbee issues.
Disclaimer: Killing power to your hubitat hub before its cleanly shutdown can cause database corruption. Use at your own risk.
I have my "Z-wave" hub on an outlet controlled by Ring.
When I want a power-off reset reboot on it, I get my second hub (the one with all the apps) involved. I shut the first hub down, wait for that to happen (or at least a long enough wait time), have the second hub ping Ring, through the community driver, to kill power, wait, then turn it back on.
This can be triggered by a Z-wave failure--or my clicking a dashboard button.
However, I thought the need to power off after such a failure was fixed recently????
Mode not Disabled? Is that a custom mode you use to disable RM?
Location Event: zwaveCrashed - is that the status that happens after the cloud backup failure?
Love the Zigbee triggers - I can use this for my Zigbee/automation hub.
I don’t use HA (probably should), but thinking I can use IFTTT to act as middleman between HE and Alexa maybe. Or does anyone know anything about Alexa automations directly?
Yes, it’s a design hold over from smartthings (PTSD) to make it simple to disable all automations when their cloud glitched. I’ve never needed to use it with hubitat except when i had eye surgery and i needed to keep it totally dark.
I honestly dont know, i dont use cloud backup. I would check your hub logs to see to verify if the rule needs any adjustments to fit your use case. So far it fits all the hub reboot use cases i’ve run into. It’s why i have 3 different ways of detecting issues with the zigbee radio .
Luckily that rule hasn’t had any updates or use in a very long time.
I have this sometimes. I just run the backup again manually and it is always successful. Never reboot or unplug hub. Don't have any problems with my zwave network. Any reason I should reboot and unplug?
Too frequent cloud backups are probably not good either. There are only a small number stored, and by the time you notice and are prepared to do something about it, maybe all you have are crap backups.
Just noticed the crash in my logs around the same time as the scheduled weekly backup. Asking again, should I reboot and unplug? Zwave works and if I just rerun the backup it is always successful.
All my automations are in Node-RED and I have a two step process.
Send an alert that Cloud backup failed.
If Z-Wave is unstable, I flip a virtual switch that triggers an automation to shut down the hub, cut power to a KASA plug that the hub is plugged into, wait five minutes and then turn on the KASA plug. The KASA integration from Node-RED is local (uses the IP address and accesses the local API).
Had my first cloudBackup: Failed steps: Z-Wave and the message recommended powering off the hub for 30 sec and trying again. This is my updated automation that notifies me and shuts down hubitat. No changes were made on the Home Assistant side for the power off/on.
Is there an api endpoint that can be used to “manually” trigger the cloud backup?
.
