My HUB after running a long time (weeks w/out a power cycle) just slows to such a grind that the ADMIN portal is about useless (clicking anything is like waiting 30-60 seconds) and events become more delayed than normal. I wish I was joking, but the only way I have delt with this is to have my Smarthings HUB cycle my HE unit via a zigbee switch I have.
Any suggestions to figure out why this thing slowly grinds itself away to the point of being useless?
I have about 30 zigbee devices and about 20 zwave devices. A small handfull of virutals and about 10 apps installed.
Sorry for the troubles. Can you list the custom apps and/or drivers you are running? As first step, I suggest disabling any custom code to see if the problem goes away. If it does, then you can add the custom code one by one to identify which one is the problem.
Just curious. How are you rebooting the HE hub properly this way? Is the zigbee switch just cutting and restoring power to the HE hub? If so, that's not good. You should be issuing a reboot from the settings page.
The OP states he's using a zigbee switch to recycle his hub, which is mainly what I questioned. Granted, a power disconnect may be the only way to reboot in the OP's situation, but doing so repeatedly cannot be good for the hub's database.
Well, you don't have much of a choice if the hub is locked up now do you? And I hate to say it but shortly after getting really slow, eventually it will lock up. Yes, it's not ideal. But what's the alternative, not using Hubitat anymore? LOL
BI Control
BigTalker2-Parent
Button Controllers
Chromecast Integration (beta)
Cobra Apps
Message Central
One To Many
Speaker Central
Device Watchdog
Ecobee Integration
Google Home
Groups and Scenes
Hub Link
Hubitat® Dashboard
Hubitat® Safety Monitor
Hubitat® Simple Lighting
Hue Bridge Integration
Maker API
Mode Manager
Motion Lighting Apps
MQTT Bridge
Notifier Plus
Rule Machine
Maybe I'm being a bit accessive wtih them, but they all have a home. I would suspect the busiest of them is the MQTT bridge and I have it rebroadcast all my devices to my MQTT server (there are some things I still do from HomeAssistant/Node-Red) that I can't do here.
I guess I'll drop the stuff I don't really need for now and pay attention to it.
Maybe for the interim I'll try to proactively do a reboot on it. It runs fine for weeks as it slowly gets slower and slower. I suspect a proactive bi-weekly /reboot might be a good option for now until I figure it out. I certainly like that better than having my ST hub recycle it.
I have not had reports of any Cobra Apps causing slowdown problems.
However, I do notice that you are using both Message Central and Big Talker.
Whilst there should be no problem with this (and there should be no compatibility issues between them) They are both very big apps!
Just something to be aware of.
I will disable Watchdog as that is one of those things I installed when "going crazy" when I first got the Hub. I also don't use Big Talker so I'll remove that as well and see how it goes. Thanks
Are there other reports of Device Watchdog causing slowdowns? I have it on two hubs and only notice the occasional slight delay on one (intentionally all stock apps except that, HubConnect, and my lighting apps--my intent being to use this hub only for lights to minimize response time.) If so, I wonder if I could make it work through HubConnect on my "coordinator" hub, which I care a bit less about--I was just thinking the check-ins and whatnot wouldn't communicate well between hubs. Hmm...
I have Device Watchdog installed and have not had any slowdowns recently.
My hub did slow down a couple of weeks ago but as I only use DM to give me reports on demand, turn switch on to update latest info, so I don't see how it could have been DM.
My device was not reponding today, but it was just the network port. Everything else, not via network cable, was working. One for the developers, but I'm not sure who to tag.
I run a ubiquiti network, and nothing else has network problems.
Edit: weirdly, the websocket and logsockets were still working, even when the network stuff was not (such as connection to wifi yeelight, or the hubitat webpage, or the dashboards).
I received a 4th hub on Saturday and decided to “video” the registration process, which went smoothly til firmware download, which hung at 40%. After waiting 30 minutes I did a ‘resetHub’ and started again. Appeared to work and I moved on to installing and ‘video capturing’ a HubConnect installation, which went swimmingly til it came time to get the remote hub to connect.
No network connection (timeout) from the remote hub. Ping worked, and the browser on the original connection, (registration and HubConnect install, including import) continued to work. That one connected pair, nothing else could see the hub in any way (except ping).
That sounds similar to what you described, @mike and my solution was to browse to settings and click the Reboot button. When it came back, everything was perfect. Tagging @chuck.schwer and @bobbyD if only to be a nagging doubt