"Hub has a Static IP address and DHCP reservation and is always updated as soon as releases are made available"
I have been a user of hubitat for several years now, about 6 months ago my main C7 started freezing up and would still display a green light and on the logs would show that it was connecting to zwave and zigbee stuff although it was completely unresponsive to wifi and ethernet commands. I could reset it by going to the ipaddress:8080 page and reboot. It would be fine for another few days / hours then lock up again?
I checked the device and app stats and even though there was nothing obvious there I offloaded the highest usage apps and devices to my second C7 hub (that does not experience the same problems.)
Because I am using a moderate amount of apps and devices I assumed that maybe it was a speed / database issue and upgraded to the C8 pro.
This worked fine for about a week or 2 then it too started experiencing weird freeze behaviour? I have been managing this by monitoring the HE’s database size and running a rule to reboot the habitat when the DB is greater than 20MB for an hour (as my DB used to sit around 10MB quite comfortably, although after recently installing the Sonos app it has been a little less predictable.)
Today the hub froze again and would not run rules correctly I eventually had to reboot it by plugging and unplugging as I could not access the interface at all through the ethernet although the light was still green and the dashboard would reset and appear to function although the ethernet devices were not able to be controlled?
(I acknowledge that this post may appear to be a little disjointed but I am not entirely sure what else to look for, TLDR: HEC7 and C8 pro freezing with plenty of free memory and storage, and nothing obvious in logs)
Please help, I have also run commands to reduce the amount of events and states to 11 for all devices on HE. I have also spent quite a few hours trolling through this community trying to find any advice on what I can do. It's quite frustrating. I am quite sure it must be a "me" problem otherwise this community would be blowing up.
Cheers
Duncan
db algorithm changed you should not mess with that.. what do you mean by reset..
try a reboot with db restore as you have corrupted the database.. are you using the stock power adapter?
1 Like
Somewhere recently, possibly 2.3.9, upgraded the database engine and the DB size will grow more before it gets auto-compacted by the hub. This is normal and nothing to worry about unless you start seeing database size alerts in the UI. So having a reboot at just 20Mb is probably rebooting the hub multiple times I day. I would turn that off totally, and then do a reboot > advanced > rebuild database.
Here is my DB size over 7 days, on both an active hub and mostly idle hub (for testing). Hub is not rebooted on any sort of schedule only if updates needed. So many of those peaks and drops are not from a reboot but routine compactions done by the platform.
1 Like
I had turned off the reboot over 20MB for greater than an hour rule after I ran the commands to globally reduce the states and events for all the devices to 11. The hub ethernet froze after a couple of days so I re-instated the rule as it keeps my hub functioning and only runs every couple of days. I was aware that the hub resized every night and that 20MB was by no means a large database hence my frustration. I have been thinking this morning that maybe its not the hubitat but the ethernet, as I use a ASUS XT12 network which is super stable but was using a static IP address with a default DNS of 8.8.8.8. I changed these this morning to the primary and secondary DNS addresses of my ISP as this has proven problematic with other devices like iphones when setting static IP addresses and changing to the ISP DNS addresses has rectified that issue.
Hi Kahn, I havent changed the power adapter, I have been using the same one for years for both the C8 Pro and the C7, the C8 pro does not come with an adapter? by Reset I mean run a rule in rule machine that monitors DB size and if it is greater than 20MB for an hour then reboot the hub.
thanks for your reply
It should, unless you are not in the USA possibly some other countries it does not come with one. I have heard on here the C8-Pro comes with a 2A so if you were using a supply rated for less than 2A then you should try switching to a 2A or higher for the time being to rule that out.
Try switching to DHCP and set a reserved IP on the router. The hub will get the DNS and everything else it needs via DHCP. The DHCP option is the preferred way to connect your hub, the static option for whatever reason has caused random problems in the past. So it would be good to rule out the static IP as the cause if it keeps happening on DHCP.
3 Likes
I am in Australia, and the power supply I have is rated up to 2.1A.
I am concerned about the spike in processor CPU pct as this seems to pre-empt the freeze behaviour. I will keep an eye on it after the DNS changes I made this morning, If / when it happens again I will try a new power supply, If it continues to happen I will try the DHCP thing. Its just weird that it was rock solid for 2 years then both the C7 and the C8 Pro have the same issue. with the same power supply and the same Network configs...
I just wish there was a log or something that could tell me why the processor was having a coniption.... sigh
I will repond after I have tried the above
thanks for the response
Duncan
Oh yeah I see the CPU spike now, and you are sure that was before the lockup and not after rebooting? A CPU spike right after reboot is normal.
Looks like the temp spiked way up right with it, suggesting something went haywire and the CPU really was under heavy load until it overheated and locked up.
Do you have any user created integrations or drivers that use LAN/IP connections (if so which ones)?
1 Like
It deffo happens before the freeze up, I use a few user integrations, When I think about it the one I potentially added around that time was the shelly plus from ShellyUSA. I only have one device on that particular app. I just checked the logs for that and its being a bit demanding. I have updated the firmware and dropped its refresh rate from every minute to "manually or polling only"
I can provide a complete list of my custom apps if it is something that will help with resolving this although with the timeline I have for how long this has been occurring I suspect its going to be a short list within the venn diagram for "when did this start happening" and "what did I add about then"
1 Like
Thats a good start. TCP/IP integrations and drivers are the first thing to look at. Sometimes if a device/service goes offline they will keep trying to open connections and it escalates quickly.
2 Likes
