In the last two days I’ve added a Lutron Pro hub and a single pico remote. When I added the hub, my Hubitat crashed. I had to pull power to get it to do anything. After Hubitat rebooted I was able to configure the pico to work with some lights without issue.
Twice since then my Hubitat has crashed requiring me to pull the power plug.
It’s possible that this is actually unrelated to the Lutron integration because I believe a new software version of Hubitat came out recently — though it was one of those versions which didn’t come with any release notes so I have no idea what the changes were.
It could be a couple of things. Sounds mostly like a corrupt database. Manually back up and download that back up (this will clean the database) Check your z-wave table to make sure there are no ghosts. After confirming there are no ghosts, do a SOFT reset (do not do a hard!) After the soft reset is complete, upload the backup you created back to the hubitat.
I don't see any relevant errors in the log, but a whole bunch of devices have some logging turned on so there is a lot of noise. I'm not sure what the retention policy on log messages is. I disabled the log messages on a bunch of zigbee devices (the descriptiveText logs seem to be enabled by default) which should reduce traffic in the logs, but it's a waiting game until the hub crashes I guess. Hopefully the log retention persists across reboots, and hopefully it's based on a message count rather than simply age.
I searched for "Exception" which is the only thing I could think to search for, and found a couple entries from the ecobee integration. I have no idea if these errors are new or not, I don't spend much time watching the logs:
app:212 2021-03-28 23:58:33.592 errororg.codehaus.groovy.runtime.InvokerInvocationException: java.net.SocketTimeoutException: Read timed out on line 464 (getCurrentFurnaceStatus)
app:212 2021-03-28 23:53:29.322 errorUnhandled exception java.net.SocketTimeoutException: Read timed out in ecobee polling pollAttempt:1, isThermostatPolled:false, isSwitchesPolled:true
I should probably state that this is a C5 hub running hub-2.2.6.130. I'll add that to the original post as well.
I haven't done anything to add or remove a ZWave device in about 8 weeks, but I suppose it's possible it didn't go well. At the time I was excluding a couple devices which are only in use during the holidays. Throughout most of the year I have 14 ZWave devices, and during the holidays 16.
How would I know if there's a ghost in the machine?
With a C-5, only by using an external secondary controller - like z-wave toolbox, or PC Controller. Since this process just started after you added a Lutron integration, I would first recommend checking whether it is related to adding that integration.
