I have recently moved my hub to a c-8. It went well. I kept the c-7 and set up hub mesh. I moved devices that communicate with the Internet to the c-7. Still all is good. That was about a month ago. Starting this week I noticed the web interface is becoming unresponsive. I can wait minutes to get a response and sometimes it fails to return before safari times out. I have had to shutdown the hub several time to get it back operational. The diagnostic interface also becomes unresponsive when the hub is unresponsive.I have had to pull the power on the hut to get it back working twice. Rules still seem to respond the same as before.
I have many devices and rules. It would be almost impossible to recreate as it was built over many months. I have a subscription and have cloud backup.
I have reviewed logs, device and app stats and I don’t see anything that stands out. In reviewing posts in the community, I may have a corrupted database. I don’t know how to tell.
I only use zwave and have the zigbee radios turned off. I factory reset the c-7 which I assumed that reset the zwave radio. I have no zwave devices on the c-7. I moved the Google Nest, Amazon, weather underground, Twillo, and Tesla devices to the c-7 as to minimize their impact on z-wave responsiveness.
I have 126 z-wave devices on the c-8. With a zwave stick I deleted any ghost devices. About 80% of my devices have a direct route to the hub.
In the past, I have had problems with some zooz battery devices generating network storms reporting battery levels. I saw the storms in the logs. I am in the process of updating firmware on the offending devices which is not easy and going slowly.
Can you post the c8 details page normally instead of the z-wave mesh details app? Makes it easier for me to read. Also on your c7 where it says RESET is what you want to do for both zigbee and z-wave so they don't interfere with the c8
How is the best way to capture the z-wave detail page as a picture as mine is many pages long and screen capture only gets one page at a time? The movie above can be slowly scrolled to see all the items.
Overall your mesh looks good (but the hops on that laundry sensor! lol) I see several devices that haven't checked in (maybe activate those just to make sure they're working).
I would next start checking logs for chatty device and your event logs around the time of the lock up to see if you can see anything off. Also maybe install @thebearmay 's hub information driver and write a rule that will tell you that the hub dropped below 120k memory and to reboot if it does. (this would also be a good indicator that a memory leak may be causing it). Also check the time the hub is going of line (after you bring the hub hub, look at the logs./event logs and see what was last reported)
as in a notification. If you were home you could maybe see things happening in close to real time.
It's like you have to be Sherlock Holmes to find the culprit. Needle in the haystack.
It started a week ago. Do logs still exist from back then?
It sounds like it's repeatable. Weirdly, that might be in your favor.
Would doing a soft reset with a subsequent restore from a backup made shortly before do any harm? I do it all the time to make my logs look good and never had a problem. I think it helped with something or other in the past as well. The backup/restore process I think cleans 'stuff' up to a degree.
I installed the hub monitor app on the c-7 to check both hubs and notify my on my twillo account. I created a dashboard with hub information driver attributes. Hopefully i can see what happening when the ui slows down.
I found the hub unresponsive again. I started getting messages from the monitor saying the the hub was low on memory (185548). When I tried to reboot, the interface timed out. This indicates a memory leak. Any suggestions on how to isolate it?
I came to the forum about to start a new topic, but I seem to be having the same issue. A couple of times in the last week, and now two nights in a row, my C8 hub has frozen up. Different times, although in both cases the last entry before the freeze is a battery info entry for my front door lock (Zwave, U-Teq using Generic Driver, battery at 100%). The battery reporting seems to be unnecessarily frequent (every 5 minutes) but not sure that could be a cause? I've dowloaded the 'hub information driver' and I'll see if that gives me any more info.
I have not, but will note for next time. Trying to go to just my regular hub IP after it's locked up and there's no response. Wondering if a soft reset is the way to go?
