The number of events stored per event type or attribute, and typically devices have multiple attributes and there can be many event types... So even thought the numbers are set to a low level, the number of events stored can go up rather quickly.
Ok, thanks for that!
That top device has 13 different event types and event history size is set to 50...so that should be a total of 650 events max? Or is there more to it than that?
As I understand it, that is how it works.
Not necessary, just that you may want to pay attention to it if something is not working.
Cleanup runs once an hour, so 650 would be the observed minimum.
If max states and events is 10 or less, it will truncate history on every save, keeping the exact number around. There is a performance penalty.
Very much agreed. There's got to be a reason to have this reporting, otherwise it should be turned off. Just because one can, doesn't mean one should 
5 Likes
The worst offender is the Aeotec 6 zwave plug. Also, my Hue Groups. I set the max for everything at 15 and 15 like 10 days ago, however my database has now grown to over 700!!!
OK, it's still getting worse.... I've set everything to 10, but the database is still growing. Now I'm at over 1100....... Anyone have an idea what's going on?
Go to you logs, Device Stats tab, check the Device Events Historybox - this will add a column with the number of events per device. Sort this column so the largest numbers are on top. Do the same thing for Driver state size.
Look at the devices with the highest counts in both columns to see if there are further adjustments that can be made to lower the counts.
On a side note the edit button is too close to the delete button in this forum, and I, for one, could use an undelete option....
4 Likes
Hm. This is what I've done. My Aeotec 6 smart plug is causing a massive amount. Can I restrict it to even less than 10 and 10 though?
I have some (i.e. Weather) set at 1 and 1 because I either don't care about past events or I'm pulling them off into an external DB for other analysis. May also want to see if some of the events can be throttled back on the sensor, i.e. do you really need the UV or Lux every time it changes or is once every 15min enough?
1 Like
OK thanks, that's helpful - I'll try and look at that! I find that Hue devices, especially Hue groups, also take up lots of storage... Would you recommend pairing those back to? Is this just about how much historical data is stored?
Also - could I just plug in a USB stick and solve the storage issue that way?
I would try adjusting the device’s settings to restrict the production of so much data. This will also help reduce the traffic on the Z-wave network.
2 Likes
USB stick won't help, but look at your use case for each device/group. How far back do you need/want data? If you only want/need current data drop it down to something low, if you want some history a little higher setting, if you're trying to debug a problem you may want a temporarily higher number.
1 Like
I've now lowered it to 5 and 5 globally, and set the Aeotec plug to 3 and 3. Currently the database is at over 1200. Let's see whether it starts going down.... 
1 Like
That's not addressing the real problem, which is the mesh spamming by the Aeotec plug. Set the Aeotec plug to report less often, with wider thresholds.
I don't know how?
Should be one or more settings on the device driver itself (generally under preferences).
OK, I think I found it... Let's see whether this helps. Thanks for all the assistance!
I don't have this device, but the manual seems to indicate that it's done by parameters 0x5B and 0x5C.
https://aeotec.freshdesk.com/helpdesk/attachments/6102433596
1 Like
None of this has helped, I'm afraid. It's still growing. I think I'm going to choose to ignore this for now and hope that things keep working. I find it so strange that Hubitat would just post such an unhelpful alert and then not enable users to actually do anything about it.
