Feb-02-2025 I updated my hub firmware to 2.4.0.151. Since then I have multiple rules that are not being triggered when a variable changes using either "increase" or "Changed". I've tried to clone the rule and re-create from scratch and no change. If I go into the rule and tell it to Run Actions it runs and executes as expected(and generates logs in the device). I can see the variable in the rule definition changing, but that doesn't run the actions. The logs for the devices prior to the firmware update show that the rule was triggering the device, but after the firmware it only shows it triggering the device when I manually tell it to Run Actions manually.
Is the only way to roll back the firmware from a backup? My backups on the cloud do not go back far enough to try to re-load.
Restoring from a backup does not change the firmware revision. If you just want to revert firmware, go to the Diagnostic Tool (port 8081 at your Hub IP) and choose a firmware version for reversion.
For what it's worth, I had this happen too just last week. No matter how many times I recreated the trigger, the correct variable change would not trigger the rule even though it had for years without issue. Cloning the rule has the same result. However, recreating an identical rule from scratch totally fixed it. Very strange.
A while back I did go back to a 2.3.9.201 and Hub has been responding as it should to variables. I'm surprised it's taken Hubitat so long to fix this bug or release another version. Any thoughts if soemthing is going on at Hubitat? Seems different than responses I've seen in the past. Thanks, BRad
Yes. The next release has major changes to the platform and the zwave subsystem. It is under active beta-test.
Edit - this is the largest platform beta-test in recent times. There are many exciting platform changes. The changes to the zwave subsystem are also fundamental, and address long-standing issues in software from SiLabs.
Yeah agree itās been awhile. Unusually so. That said, there are bigger more important changes afoot and there are plenty of work around for this little issue (once known).