I have a couple devices that seem to be creating an excessive number of Hub Actions for the amount of time they are being used for anything. One device in particular, "Kasa5" pictured below, is installed in Hubitat but is currently not in use by any rules or automations. The device just sits on my desk until I come up with a new purpose for it. However, even though it just sits there and hasn't been touched in weeks, it seems to be relatively active within Hubitat for some reason. Why would this inactive device be generating so many Hub Actions?
The device is also relatively near the top in terms of % of busy, which makes no sense considering there are literally no apps, rules, autiomations, or anything using this device at all.
It should be noted that this device, and the other two below it in the image which are also producing a lot of Hub Action are all Kasa devices using Hubitat's build in Kasa integration.
Kasa relies on polling to receive changes made from outside the hub. That's going to happen whether or not the device is online (keeping it off might actually cause errors and make things worse, though probably not enough to matter with reasonable polling intervals).
Are you noticing problems or is this just curiosity? If no problems, I wouldn't worry about this. Polling causes activity on the hub regardless of what the device may have done (or not).
No specific issues. I am not totally sure what "Hub Actions" even are... I just thought it off that a device with no automations at all that isn't being used at all (but is powered on) would be producing the largest number of hub Actions. Then I noticed that Kasa devices in general tend to produce the most Hub Actions - and generally the least used Kasa devices seem to have the highest Hub Actions counts....
They are normally Zigbee, Z-Wave, or LAN commands sent from a driver, here most likely a poll. (If you catch it during, the Hub Actions tab in Logs will tell you more, but they normally run--and disappear, then--pretty fast.)
I'd expect most to be about the same, but it's possible there are retries or something for your offline devices that might be causing a bit more for them (just a guess, haven't looked at the code to see how it handes that case).
1 Like
Looks like the polling frequency for those devices was set to 30seconds or 1 minute. All others were at 30 minutes.
1 Like
