I just got a notification saying one (some?) of my rules are creating excess notifications. The notification cleared before I had a chance to look at it closer.
Is there a way to get that notification back, or find what rule is creating the excessive events (although I have a pretty good idea)? And how many events are classified as "excessive", and what is the downside to excessive events?
I believe the default threshold is 300. You can look on an individual device on the Device Settings tab to see (and set) the threshold. But that warning is for a reason...
You will slow the hub or cause a crash.
Go to Logs tab, and see if there are any warnings or errors in either current or history logs. Handle those first.
After that, go to App Stats in Device Stats in the Logs tab. Open the settings just above the statistics in both of those, and check all the options.
Those App Stats and Device Stats should tell you if something is chewing excessive processor time. In this example, my overall usage is very low at 3 minutes for a whole day of uptime. And my worst app is barely working the hub at all.
In case it's helpful to know, this was almost certainly a device, not a rule (or any app), and it was probably "excessive events," with the steps above probably being the best way to troubleshoot. If there are any current alerts, the http://hubitat.local/alerts page (replace "hubitat.local" with your hub's actual IP address if this doesn't work) will show them--the same place you'd get to if you clicked the alert in the notification when you got it--but if they're gone, they won't be there, so...might not be of much help now.
yes, you are right. The warning popped back up and it was a device, and it was excessive events.
Thanks. I see that this was a device that I had set to poll a bunch of different attributes every minute, and then I had a rule that ran several different actions based on those attributes.
I have now changed the device to poll every 5 minutes, and that seems to have calmed things down. However, this is also a time of day where those attributes aren't firing off a bunch of different rules, so it may be too early to tell if it is OK?
The rule (an app) isn't related to the message you received--it's just saying that the device is generating lots of events. Events are generally any time an attribute value changes. I'm not sure what specific device you're working with, but with a dimmer, things like level are attributes and have a value from 1-99 or 1-100 representing the brightness. A dimmer also generally has a switch attribute with a value of on or off. The total number of events generated by the device is all this message is complainng about, so the more attributes and the more changes/events you have, the more likely you are to get this warning,
It should also be noted that this is just a warning and not necessarily a problem. A lot of devices doing this could be a problem, as it could make your databae large, and large databases can cause slowdowns. Something that truly hogs all the hub resouses could cause even worse problems, but a device that's just polling every few minutes probably isn't one of those. But it would still make sense to reduce the polling frequency, number of attributes, or anything else that might cut down on this if you don't really need the information. If you do and don't want to get this warning again, you could also increase the "too many events" warning threshold for this device (I also think this is 300 by default--just a reasonable value chosen by staff, nothing else magic about it).
In any case, that's where you'd look--not the rule. But if you have a rule--or any app--triggered by these changes and they happen "excessively" frequently, or if you have an "excessive" number of apps subscribed to any/all of these events--then that's also more work for your hub. There are some things you could use to reduce that work, too (e.g., if it's a rule and you only care about your trigger event during certain times of the day, you could add a required expression for that timeframe). App stats (and device stats for the above) can help you figure out what's taking time here if you do have problems, but it seems less likely--and it's not the source of the warning. I don't think there are any app-related warnings you'd get like this, just this kind of troubleshooting you can do if you think there is a problem.