My Hubitat devices briefly (for about 4 seconds) go offline on an hourly basis. At the first minute of every hour (ie. 9:01 am), all my Hubitat devices show as unavailable for 4 seconds before showing available again. Example below:
This makes it very difficult to use automations and conditions based on state duration. I know I can create template sensors that ignore changes to “unavailable” and use triggers that specify the “from” and “to”, but it makes everything so much more complicated, especially since I have around 70 devices coming from Hubitat. This applies to my Z-Wave, Zigbee, and virtual devices.
I can’t figure out if this is coming from Hubitat, the Maker API feature, or the hubitat component in Home Assistant and if there is any way to block this hourly reset. Is this typical for this integration? Does anyone else experience this?
Found these.
Generally, though, as default it appears they both get their time from the the DHCP server, so make sure they are both using the same one. I suggest using the local router as DHCP, then have the local router sync with a time server and external DHCP.
Ugh, feel like an idiot but I finally figured out what was happening here and it had nothing to with time servers… a while back, I was using a bunch of Tuya light bulbs that were flaky at best. Since they kept becoming unavailable several times a day, I made an hourly automation to reload the Tuya configuration using the homeassistant.reload_config_entry service call. I had it run 1 minute after the hour so that the lights wouldn’t miss any lighting automations which often run on the hour or half hour. My service call wasn’t to the config itself but rather to the tuya entities, which makes the call reload any configs tied to those entity names.
When I finally got fed up with trying make my old tuya stuff work, I replaced with zigbee/zwave connected through Hubitat but used the same entity names so I wouldn’t have to edit any affected automations/dashboards. So now my old “restart Tuya hourly” automation was calling those entity names and also restarting Hubitat hourly.
Feel like a moron that it took me this long to figure out. I was really overthinking what was causing this.
