But devices were transferred? And apps were? Rules?
Nothing is showing in logs? (errors or warnings)
App Stats and Device Stats in the logs tab are not showing any excessive memory usage?
How much does this concierge service cost?
6 Likes
Lets see.
- Plane ticket from HQ.
- Per Diem for meals and incidentals.
- Hotel room for at least one night.
- Hourly rate charge, minimum of 4 hours.
- Rental car.
- Materials as required.
- Taxes and fees.
I say probably $3K per visit, but that is a wild guess.
3 Likes
@neonturbo No - all devices were factory reset and added to the C7 as if they were brand new.
As far as errors and warnings, nothing.
As for App Stats and Device Stats what would be considered excessive?
Sorry but that's completely ridiculous to suggest that you'd send a technician on-site. This isn't 1995!
One major problem with this platform (and I'm not the first to ask) is that aside from the device event log, Hubitat does not have a proper system logging system. I'm talking from the OS up that articulates everything that the hub is doing or has done. It would be really nice to be able to point the Hubitat to my syslog server which would likely capture the problem or point us in the right direction.
Syslog is kind of a rudimentary function for most networked products and I can't believe Hubitat still hasn't bothered to implement this!!
If I could capture that data, it would be so easy for me to package up the logs and send it to you for analysis / replay. Problem solved your issue appears to be caused by (example) X device or rulemachine has a bug and whenever Y occurs, your rule is causing it to freeze.
For this I'd be happy to pay an annual support fee of a few hundred dollars so someone could provide guidance.
3 Likes
Maybe transferred wasn't the right term. But you did bring/add/include/pair (or whatever term you would like to use) all the devices from one hub to another, correct? If you had a bad device slamming the network or had an app that was runaway on one hub that you are also using on the second hub, you brought the problem with you no matter what means you used to transfer the device.
To me, it stands to reason that it isn't the hub but rather something common to both hubs causing an issue. That is why I asked whether you were using the same set of devices, apps, and so on. Likewise your home network or UPS or anything like that would be common to both "failed" hubs.
That was supposed to be a lighthearted if not silly reply to the suggestion that they would actually send one of the staff to your home. Of course they won't do that. It is funny to envision BobbyD coming to your house in a brown and green Hubitat uniform! Instead of Geek Squad they could be Hub Pals, or Hubitat Syndicate or something like that.
I am not sure there is such a thing. Maybe staff could verify. But what would you do with that info anyway? It isn't like us end users can fix anything in the Hubitat OS. And most of us aren't trained with or familiar with the underlying OS, so the best we could maybe do is pick out an error message. Ultimately if there is an issue, the engineering team has to patch things and send out an updated firmware.
3 Likes
They very much do and Bobby said as much above. These logs just aren't exposed to end users (I presume do to them containing intellectual property or as a "security through obfuscation approach"). Given that Bobby did check those logs, did not find a problem, and sent you a replacement hub that's exhibiting the same issues, then I'm with @neonturbo that something common to both hubs, but external to the hub OS, is your culprit.
The syslog driver that rlithgow linked to will send out the device/app logs to an external server via syslog.
2 Likes
Yup. The one thing that Hubitat needs to fix is the hub’s network interface freezing when it encounters a jumbo ethernet frame.
While there is a workaround (placing the hub on a network segment with no jumbo frames), a better permanent solution would be to fix it at the OS level.
@kj1 - my best guess remains that there are still devices on your hub’s (V)LAN that are sending JF.
4 Likes
@FriedCheese2006 - Bobby never checked my logs (that was someone else in the thread that had an issue with their AC adapter if I read correctly) second, the replacement hub I paid for out of pocket only to find out there was nothing wrong with my C5.
The git link is broken but I'll search it out over the weekend and implement it. Thanks @rlithgow1 .
I misinterpreted you comment @neonturbo
As for jumbo frames, its disabled across all vlans.
Oh and I forget who mentioned my UPS as a possible cause, we can rule that out because when I received the C7, i've had it plugged into a regular wall outlet and experienced the lockup
Unless someone can actually review my logs (please) I think the only option is process of elimination. Might try move 5 devices at a time over to the old C5 to see if that stops the C7 from freezing and gets the C5 to freeze.
I'd generalize that statement to 'resiliency'. Network resiliency. Database resiliency. Z-Wave resilliency. And on.
2 Likes
My mistake. I missed that Bobby attempted previously but couldn't get to the new hub due to cloud connectivity issues and that you replaced the hub yourself.
1 Like
Look lower in the thread, they post the raw driver. That said I'm inclined to think something is up with your network and the jumbo frames aspect fits. That said start simply. Isolate a small switch and connect directly to your router and see what happens
1 Like
No worries.
While I doubt my LAN is to blame, I'm going to rule that out first. I forget who made the suggestion but I've decided to disconnect the C7 from the LAN over the weekend as I'm heading out of town for a couple days. (realize that I shouldn't advertise on-line however others who are not technical will be home)
I have a number of automations set up, door opens, lights go on, lights on at dusk, lights off at sunrise, etc.
Will see what I come back to.
2 Likes
Sorry to interject into your discussion. I have the same problem. I found this discussion after posting here (https://community.hubitat.com/t/hub-stops-working/110616) unanswered.
I don't have the same expertise as you guys but I was able to follow along somewhat but unable to resolve the issue. I lose device control after several minutes after a reboot.
So, where I last left off…. Last Thursday I disconnected the hub from the network (physically) in an attempt to rule out the network as the root cause before heading off for some indulgence.
To my surprise, I arrived home last night to find the hub fully functional. All my automations responding normally.
As mentioned, my network consists entirely of UniFi products. Jumbo Frames is turned off so any ideas on what’s causing this or how to trouble shoot?
What are your logs showing during the time of the freeze?
The hub uses ARP, IMCP Ping, DNS, mDNS and various other services. If you have anything like that blocked and the hub keeps pounding it, maybe causing it to spiral out of control? I have seen other posts from people complaining about constant pings and it was mostly because they are blocking so much stuff the hub keeps knocking.
Also in your networking settings on the hub, there is a link speed option, whatever it is set to, try the opposite, sometimes it helps with various issues. Check the Auto reconnect check as well, I would try with it disabled and also at the highest value, see if either of those helps. The auto-reconnect was one of the things causing constant pings depending on how it was set.
Maybe someone with Unifi equipment has some ideas on settings on the router itself to check.
If using the standard Unifi firewall, there would have to be rules set up specifically to drop access based on these specific protocols. You would see these in the firewall rules list, likely in the internet tab (but not necessarily).
2 Likes
Within the UniFi setup there are global and local settings for jumbo frames. Each UniFi switch, including the UDM routers, has the ability to use global or local settings.
On the UniFi network setup make sure multicast DNS is enabled. Check the port settings to make sure things look OK there.
On the Hubitat networking side everything set to default works with UniFi equipment. No need for DNS override. No need for static IP (DHCP reservation should be set in the router/DHCP server). Ethernet speed works with auto-negotiate. Auto reconnect should not be needed.
If you have a bunch of VLANs and/or firewall rules then you can post them for someone (not me) to answer.
2 Likes
