Hub lockups?

Bought one! :smirk: Now I will have to learn how to connect it all up.

Yep.
Dev = devices
App = App

1 Like

Honestly I would not bring anything new into the mix while you haven't figured out what's causing the current issue. I'm just about to remove Homebridge this evening as a troubleshooting step.

Why not just disable it instead?

Can I disable the app? I remember seeing that was going to be added, but I'm not sure how to get there. If I just shut down the Node.js instance, it generates errors on HE because it cannot reach the host.

[Edit] NM. I just found it. Thanks

Well I thought that disabling and uninstalling all 3rd party apps was going to fix the hub lockups but it hasn't. It went about 6 days, which is better than locking up every day. I am running only HE apps right now to try to determine the issue. The only thing that I setup on Friday was Other Hub Event Pusher to send the status of two virtual switches to ST. Is this causing the hub to lockup?

Do you have the latest firmware loaded? Only checking because they made changes to memory to help with lockup’s.

I must admit the number of lockup posts have decreased since. Or they just aren’t posting.

Keep in mind 3rd party drivers can also cause issues.

1 Like

I do have the latest firmware loaded. The only 3rd party drivers and apps are Konnected and SmartLife RBGW Controller. Both of these are wifi so they don't use a cloud.

I've had the same experience, even got a new hub and started over to make sure I kept things as clean as possible from the start.

No WebCore ever on this hub, but the same stability issues. It's at the point where after the reboot (via a POST to reboot) the hub freezes at 70% initializing.

The exact same thing that happened with the last one. I'm starting to thing that there's a bad interaction between one of the OOTB device handlers and one of my devices that's causing either a lot of errors or a lot of extra logging that I can't see.

The last hub got to the point where there was database corruption and issues with writing to the storage. Feels like this one is on the way there.

Have you run into the lockups on restart yet?

I was having lockups every 1 week to 2 weeks. I ended up disabling all my 3rd party apps excluding my personal ones and Konnected. Still had a lockup occur. I then disabled most of my personal apps (except Occupancy Lighting because that one is very simple and an absolute must for my wife). Since then, no lockups :crossed_fingers:.

I just re-enabled Konnected app this weekend. If my hub locks up anytime soon...I think I found the culprit. I will let you know but it might be will be a couple of weeks.

P.S. the fact that they don't use the cloud does not stop them from causing lockups. Any bad or incompatible code can cause issues.

2 Likes

Not sure how you do a POST to reboot. I can't access the hub at all when it locks up. I have to pull the plug.

Let me know how that goes. I am finding it difficult to implement a system if I can't use apps or whatever we need to run all the various devices to make our system work. Needs to be more stable.

2 Likes

I never liked this app, and it causes the hub to be slower.

I never had a problem with Konnected ( one small switch glitch, it was corrected with a driver update), I use it from ST days, then I used Hubduino when Konnected was not available here, then from day 1 of Konnected in HE I'm using it.

I believe I read somewhere this was causing some problems, besides it's wifi it doesn't means it will not lockup your hub, I had a problem with MagicHome RGBW driver making my hub slow, the controller is wifi too.

You can try just disabling apps and devices. You can backup your hub, soft reset the hub, restore the backup, so the database will restart from the backup.

What are those buttons from HE to St for?

I may be mistaken, but I believe @pcgirl mentioned in another post that she still relied on webCore in ST for some of her automations. I assume this is sending data to ST to be processed by WC.

With every update of RM we are seeing a lot less reliance on WC. The new version is probably going to be release soon and hopefully will get most people 100% onboard with RM.

I have two virtual switches statuses being sent to ST to trigger Modes and run Ecobee because I think the Ecobee app (Ecobee Suite Manager) was causing problems. I haven't read anything about problems with SmartLife RBGW Controller. I had an issue a couple of weeks ago after a hub firmware but an update was posted to SmartLife RBGW github and seems to fine after the update. Thing is that when I search my logs, there are no errors. And really, what is the point of disabling devices so that they aren't working with HE. If I can't use them then I need to find another system that I can use them on.

Yes, just for presence detection. But only because I have disabled the 3rd party apps in HE that would be doing what WC in ST is doing now. The virtual switches just change the mode so as to set the Ecobee temp.

I hear you 100%. The past few months have taken me quite a few steps backward with my automations. Unfortunately it was a necessary evil because it's the only way to narrow down the issue. I wish I just disabled everything from the get go because I wasted a lot of time removing one thing at a time and had to wait at least a week before the hub would tell me "nope, try again". Painful process, I know.

While you may have stated this, I can't be bothered to read 57 posts :smiley: ...

Have you attempt to soft reset the db? Last lock up I had I did this, and restored from a back up afterwards. No lock ups since, fingers crossed.

Is a soft reboot from within the dashboard? If so, I just did one when I updated to the latest firmware the other day.

@stephack have you been able to determine what was causing your lockups? Perhaps a post should be started to list things that DON'T work with HE.

No. Soft Reset. Using the 8081/resetHub link documented elsewhere in the forums. A soft reset I believe wipes only the db, not the radio. You can then use a back up to restore the db without having to setup all the devices again. Solves some db corruption issues, I think.