C-8 to C-8 Pro Migration Works, Yet Doesn't

🎬 Get Started Using C-8 Pro
1

I had a C-8 Hub that recently died. I opted to upgrade to the C-8 Pro. I had a migration cloud backup ready, so not worries.

I have followed the instrutions to migrate to the new C-8 Pro and everything went great, except everything seems to be in "Read Only" mode. I have an array of Z-Wave, Zigbee and Virtual devices, including switches, open/close sensors, water sensors, plus connetions to Hue, Netatmo, Ring, etc. The state of all of these devices shows exactly as they did at the backup but none of them will change.

For example, I have a virtual switch. The switch shows as "On". From the Devices screen, I switch it to "Off"... but nothing happens. It remains on. If I turn on logging, I see the log entry of the switch being turned off, soon followed by a message that tells me the switch is "on". This is true of every device - Z-Wave, Zigbee, Virtual, etc.

I have Maker API installed and it feeds the status out to the world, but again, no changes to any device state is registered.

In addition, I am using Webcore, currently installed as an app. (I haven't migrated to native, yet). Trying to start webcore just hangs the app.

I have tried a few tricks, including reboots, power cycles, etc.
I tried doing a restore from the cloud backup, then creating a local backup and doing a soft reset from the local backup, with power cycles after restore.
Nothing changes.

Platform Version: 2.3.9.160
Hardware Version: C-8 Pro
Nothing fancy in the network setup... certainly nothing that has changed since the C-8 was working and decided to bite the dust.

Any ideas of things I should be trying? It is incredibly frustrating.

2

When I migrated to the C-8 pro, the first attempt resulted in something similar, although not quite as bad. I believe my zwave devices did work, although very slow to respond. But many of my apps and device drivers were completely non-responsive.
What it turned out to be was something went awry with my scheduled jobs. When I went into Logs > Scheduled jobs, the first couple dozen jobs listed had next run times in the past. I chocked it up to waiting a few hours between taking my backup from the old hub and restoring it to my new hub. To fix it, I plugged the old hub back in, made a new backup, then immediately restored it to the new hub.
I know you cannot do the same, but I’m curious what your scheduled jobs screen looks like.

2 Likes
3

@user3786

Start with shutting down the hub and unplugging power for 2 mins. Power back up. If that doesn't resolve the issue then I would repeat the migration and see if it clears up.

4

Good thinking! I just checked and everything in the logs looks appropriate, with everything scheduled actually scheduled to happen in the future, not the past.

5

Done. Multiple times and always with the exact same results.

6

What platform version was the old hub running when you took this cloud backup? Was it also running .160?

7

Actually it was .158. At first bring up the C8 Pro upgraded itself to .160. I’ve since updated to .162 thinking that might shake things loose. No change.

Is it possible to downgrade to .158 and migrate again from the backup? It’s basically the only thing I haven’t tried up until now.

8

I did get on the diagnostic tool and managed to revert to .147. It and .160 were the only two options, presumably because .147 is what it shipped with, .160 was what it upgraded to.

So I dropped back to .147 and remigrated from my original cloud backup. I’m still seeing the same behavior - nothing can be changed. Even a simple virtual switch won’t change from ON to OFF.

10

On your old hub do you have any ghosts that could have migrated?

11

At this point, I’d probably try disabling your apps and see if devices start working again - at least the virtual ones.

What happens if you create a new virtual switch? Does it actually toggle?

12

None that i am aware of. I had done a thorough house keeping not too long ago.

13

New virtual switch created and toggled perfectly!

14

Sheesh… seems like you’re in a pickle. Doubtful it’ll do anything, but perhaps disable all your apps on the apps page, restart and see if anything comes back to life.

15

Hmmm… a glimmer of hope. It seems that by creating the new virtual switch, and perhaps toggling successfully, I have “unstuck” the rest of the system. All devices now responding, all apps successfully underway. Even Webcore appears to be functioning.

I am thrilled, yet cautious. Without understanding the root cause of what happened I worry I’m living on an unstable platform. Any thoughts or speculation on what might have happened that got me back in business?

1 Like
16

Fingers crossed… but now I do want to know what the heck went wrong.

1 Like
17

Ok - Here’s a quick update. At some point I’m thinking I’m going to need some guidance from hubitat support, because something is not quite right.

Here are the steps I took to “recover”:

  1. I restored back to .146; symptoms remained
  2. I created a virtual switch and successfully toggled that switch
  3. Everything magically started working.

After that I tested everything - MakerAPI, Webcore, etc - all seemed well.

I updated a number of apps, including Webcore. Everything is great!

THEN I updated to .162. Everything stopped working and I was back to original symptoms. I tried to create another virtual switch and toggle it. The switch created, but the state was (null) and switching it on/off did not change the state!

SO I went to diagnostic mode, fell back to .146 and when the Hubitat rebooted everything was working again as expected.

At this point I feel like I’m revlocked to .146 and have built a house on shifting sand.

It feels like some deep internals analysis is going to be required by someone with more knowledge than myself.

18

That’s too bad. I’m certainly not an expert here, but have you tried rebooting with a database rebuild?

1 Like
19

I have not but it is probably a good idea. Right now I’m trying to ensure I’ve got stability where I am and get a good, solid backup in place. At that point I’ll upgrade to .162 again and reboot with a DB build, and will report results back here.

20

Well, just did a backup and went through an upgrade to latest release, 2.3.9.164. Upgrade was successful, devices again not switchable (back to original symptoms). I rebooted with a database rebuild … sadly, no change.

Now I am STUCK on 2.3.9.146, which is the ONLY release that seems to work!

21

@bobbyD is there something in the engineering logs that might help diagnose the OP's issue?