Guys,
I had the need to reboot the hub at least once every couple of days as automations are slow to trigger after a few days.
Is this normal?
If so is there a way to automate it? For example a Cron job every X amount of days?
No!.
Something is dragging you hub down and using resources without releasing them.
The first thing I would do is disable any custom apps you are using and reboot the hub.
If this cures the problem (test a for a few days) then I would enable them one by one with a few days in between .
If one of the apps is causing the problem you can find out which by this method.
If disabling the apps does not cure the problem then I would remove any custom drivers and do the same with them.
Andy
2 Likes
Normally rebooting the hub sorts the problem for a few days.
I had a lot of debug active, which I have now disabled, as this can be a root cause. My reasoning is that the.database cannot held so much logging. Just a guess though. Let's see how it performs.
Apps I only have the following non core apps:
- Netatmo app that's no longer supported as the developer. Left Hubitat. But I need this as it's important to track Co2 levels and noise levels due to.health. in Reality I need to find someone to port as well the sister app for the health coach.
- Device Watchdog (parent and child).
Drivers that are user based just the Netatmo and the Xiaomi. None have been reported to cause this to other users.
It has to be something unique to my configuration that nobody else's have.
Nothing else comes to mind 
That's not normal.. to echo @Cobra in my experiences:
- WebCOre issues (remove immediately, does not work well with HE)
- Custom Apps spamming or otherwise causing slowdowns.
- Custom devices doing the same.
- Devices on the edge of range.
- Faulty device causing weird routing issues.
As the folks at HE keep telling me.. logs are your friend.
1 Like
That's the thing logs tell me zero. Not a single error or strange lines.
It seems everything is held like 10 secs in cache and than triggers as expected. If you didn't know the correct time of the action looking at the logs everything would seem normal.
Also to note this does not happen straight away. After a reboot it takes roughly 5 days for me to start noticing a delay.
I have now reboot and everything seems fine
Yeah that's odd. You have a lot of Zigbee devices?
Also I wonder if you could temporarily disable the Netatmo app and see if that helps..
I have Only ZigBee devices and cloud integrations and chromecast
Hows your network connection? Could the cloud integrations be part of the issue?
1 Like
Not for 5 days... We used it to monitor Co2 levels to warn us when c02is above certain limits
1Gbp download and 200mbp upload. Sure is not it 
Well how about your local area network. Maybe there is some issue there - I had a misconfigured router that was causing my sonos stuff to go crazy.
885Mb download in every room and 189 upload.
I have an Orbi mesh system!
I do too! I really like that setup!!!! Works fantastic!!!!
Well I dunno about your issue - has support come up with anything?
You might have to start simply and add things a bit at a time..
1 Like
Best thing ever for WiFi coverage. Money well spent!!
Let's hope was all the debug that I had active.
1 Like
If you figure it out - post back I would really be interested in what the issue was.
1 Like
My slow downs of the past never had too much to say in regards to the log. Kasa (TP-Link) and Wemo being removed helped.
Obviously as stated above WebCoRE removal was the first thing a while ago.
With WebCoRE it happened about every 4 days.
It was really weird.
Almost predictable.
1 Like
I've had issues a couple of times now where over time my hub slowed, automations began delaying and then pretty much stopped.
I did see an occasional error in the logs that to me seemed to indicate a memory leak or something killing the SQL engine. Reboot resolves for a short time. After completing a soft reset and restoring from backup I've been running ok for a few days now. (Although I'm patiently waiting on 2.0.5 to drop so my automations based on lock codes can begin working again.
The error I was seeing when this occurred was:
#HubVersion=2.0.4
Any performance stats we can see? Possible to SSH into Hubitat and check?
Yes, if you know the secret ID & pw Hubitat has. 
You can watch the boot on the HDMI port and plug in a USB kbd.
Otherwise, no.
I was using a similar app, Device Monitor, and noticed automations and the hub interface itself slowing down significantly over time. A reboot would make things normal, and it would take some days to slow down again.
I removed the app, and everything has been normal since.
1 Like
