C5 hub unresponsive after upgrade to 2.3.5.110

thanks @thebearmay i thought i read it correctly. shame really (for us and hubitat), this scenario would be the perfect 'gateway drug' for the Hub Protect service if it mitigated the need to spend hours opening up wall cavities and step-ladders on ceilings to re-pair all my devices. But if I have to do that anyway...

C-8 Migration would allow the zigbee radio to migrate... (question then is, is $149 enough value to not open the walls...)

BTW have you enabled the cloud and DNS checks in the driver? (Just curious if one or the other could give us a clue on what may be happening).

1 Like

it would be if i hadn't already bought a brand new C7 only a couple of months before the C8 was announced. thats a double whammy...so its double if i took that route :sob:

no i haven't. if the symptoms were "just" a loss of ethernet/IP connectivity and the hub was alive and processing mesh automations, i would definitely have done that the second you suggested it. but seeing as the whole hub goes dark i didn't think it would any value to the situation. do you agree with that thought?

Don't think it will solve your issue, but wondering if a loss of one or the other could cause a snowball effect (again a slim chance).

hmmm..well i'm going to have to hard power cycle the hub again anyway, its been down a while now. so i might do that for kicks when its back up.

This has really highlighted to me the implications of hub failure when the house HVAC and lighting systems depend on it. and its night-time in summer :slight_smile:

on that topic, why don't we have hot standby or active/active dual hub capability? i know this topic is much broader than hubitat - but if the industry is serious about automation being the default for the built environment, this capability should be the starting point in the system architecture. hardware should never be a single point of failure. anywho...back to the fun!

The zigbee synchronization counters get in the way, but:

OK this is getting beyond a joke. Original C5 still hanging. Setup new C7, restored backup and in process of migrating devices over to it.
I did OK with half a dozen or so zigbee devices just by adding the device on the new hub and pressing the reset button on the zigbee devices (apart from the nightmare of getting access to the pairing button on the damn things in some locations).
Twice now i excluded on C5/included on C7 a zwave device (well re-paired from the C7's perspective), and then the device configuration ended up being a nonsensical mish-mash of configuration from non-related other devices. i did another restore to clean up the configuration again, but the Zwave radio registration had zombie devices against it. aaagghh. reset zwave radio, restore from backup, start all over again.
on top of all this, the new C7 hub has now hung twice. w.t.f.

@support_team please help, i've been good trying to sort it myself, but this is getting out of hand.

Thanks for your feedback. It is not clear based on the details provided if you used Hub Protect to restore Z-Wave network on the C-7, or you are rebuilding both mesh networks. If you have started from scratch, be sure to check out these documents:

https://docs2.hubitat.com/en/how-to/build-a-solid-zigbee-mesh

https://docs2.hubitat.com/en/how-to/build-a-solid-zigbee-mesh

1 Like

rebuilding the mesh on the new hub. (C7)
note that this has been a rock-solid system/mesh for probably 2 years.
nothing has changed in this (or any) regard up until the C5 started hanging.
no new (or removed or moved) devices, no new IT devices on the LAN.
and regardless of that, i would expect the hub to stay alive even if the mesh was slightly suspect. do you have any troubleshooting recommendations, or are you able to look at some system logs?

"Nothing has changed" is probably the most common misconception. Changes are being made every second without users' input, as every connected device to the Hubitat Elevation hub makes changes to the hub. Connected devices can misbehave and would have negative impact on the overall hub's performance. I am not saying that your C-5 hub couldn't have developed a problem without it being triggered by any number of connected devices, but more often than not, that seems to be the case in most systems where user didn't make any physical changes to the hub.

OK. so can you recommend a non-destructive way to troubleshoot this, the logs on the system that i get access to show nothing of value (to me at least). there should be no jumbo frames on the network, i went through all network devices and actually re-disabled them.

I always start with live Logs or past Logs to see if there are any errors posted. Those should pop out immediately because they have a red tag. If you see any errors and you don't know how to resolve them, you may want to post them in here, so others can help you, including Hubitat staff. If there are no errors in your log I'd be more than happy to check the engineering logs for you to see if something else is hitting the platform. Generally the engineering logs are good to highlight hardware issues, network issues and/or power outages. Other than that user logs are far more informative.

Hi Bobby. there are no errors of note. if you could take a look that would be good, especially network-style issues would be good to confirm, especially since the C7 has shown issues too.

Sure, send me a private message along with your hub ID and I'd be more than happy to take a look.

1 Like

Still in all sorts of problems. Both hubs regularly hanging.
I went and replaced the unmanaged 1G switch my hubs are plugged into today, with a managed switch that supports port mirroring so I can run a wireshark capture to see what i can see.
in a new first, while i was recovering yet another corrupted database on the C5, after the soft reset, it was sitting there 'blank' waiting for a restore/setup when it hung again. so this tells me its nothing to do with the configuration on the hub, because it had no config and still hung.

so i think this tells me its either:

  • external (e.g. jumbo frame/broadcast/multicast from a LAN endpoint thats gone flaky)
  • hardware (but unlikely seeing as both a brand new C7 and a several-year-old-but-historically-rock-solid C5 are both showing the same symptoms) or
  • something weird on the S/W that no-one else seems to be experiencing just me.

just hope the wireshark exercise helps here. i'll be truly stumped if it doesnt...

Are you running jumbo frames on the network?

Hi there @rlithglow1 no...we had this conversation :slight_smile:
i have been running packet captures last couple of days to confirm assumptions though, and i have not seen one single jumbo frame. but these damn hubs keep hanging. :triumph:

1 Like

Outside of doing another softreset I don't know what to offer. Outside of the previous stuff discussed the only thing I can think of that does that are zooz 4-in-1's and zen-25's

1 Like

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.