Continued C8 Pro hub performance problems - new thread


I am starting a new thread / hopefully opening a support ticket on my hubitat c8 pro.

. After working ok Tuesday - Wed morning initially once again z-wave performance gone / non existant. Definitely good mesh, no ghosts, I ran z-wave logging last night I am about to feed that into AI. Hub light is on and green, but the hub / app itself was VERY slow / non responsive about an hour ago - but now the hub is accessible again (but z-wave still crappy).

AI summarized my z-wave logs as this …

Overall, the vast majority of your mesh looks beautifully healthy, but there are two notable entries to look at closely:

1. The Dynamic Routed Device: dev:712 (Highly Active) (zooz z-wave motion sensor)

On pages 1, 3, and 5, dev:712 is quite active. Looking closely at its routes, it is consistently bouncing between different multi-hop paths to get back to the hub:

  • It frequently routes via repeaters: [0x3e, 0x15] at 100 kbps with a 103ms transmission time.

  • It occasionally drops down to speed: 40 kbps via repeaters: [0x3e] with a 144ms transmission time.

  • Its entries show routeChanged: true quite a few times.

What this means: routeChanged: true isn't a failure—it means the mesh is actively doing its job. Because dev:712 is likely a bit further away from the hub (indicated by it needing 1 or 2 repeaters), the Z-Wave stack is dynamically shifting paths to find the most efficient route. Its transmission times (around 100ms–140ms) are completely fine for a routed device, so as long as it isn't lagging in real-world performance, its behavior is normal and healthy.

2. The Slow Direct Device: dev:267 (zooz zen 04 plug -aways away from hub)

On pages 2 and 4, dev:267 shows up regularly.

  • It is hitting the hub with repeaters: None (direct route).

  • However, it registers a latency of 1021ms to 1022ms and lists speed: Unknown.

What this means: When a direct device takes over a full second to communicate and shows an unknown link speed, it usually means the hub is having to re-transmit the packet multiple times before it finally gets an acknowledgment (ACK).

  • If it's a battery-powered sensor: It might be sleeping right as a message finishes, or it's an older Z-Wave non-Plus device.

  • If it's a main-powered switch/plug: It might be positioned right on the physical fringe of your hub's direct RF range, causing packet drops that require retries.

Summary Checklist

  • dev:633 Removal Success: The network is completely clear of the rapid-fire per-second clutter. Bandwidth is wide open.

  • dev:712: Standard, healthy mesh routing behavior. Nothing to worry about.

  • dev:267: This is the only device experiencing notable latency (1 second delay). If you notice this specific device ever being sluggish to respond on your dashboard, you might want to look into its placement or see if a nearby line-powered Z-Wave Plus device can act as a better repeating neighbor for it down the road.

Everything else in these logs—like dev:912, dev:62, and dev:1094—is operating flawlessly at a blazing fast 1ms direct speed!

Notthing jumps out to me in the logs.

Wednesday, I redeployed my older c7 hub. I had to add a few devices to it that I had put on the c8 pro since May … minor stuff.

The c7 is performing flawlessly …. but that is without the addition of that yale door lock and the 4 smartwings shades.

My plan is to add the yale lock tomorrow and then test for 3-4 days. If it looks good - I will do a fresh migration to the C8 Pro after factory resetting it. Then run IT for at least 3-5 days and see.

For Rick and others - I want this to remain open as I still feel there is SOMETHING wrong with that c8 pro hub. A reset may cure it - we shall see.


For those of us that didn’t see the previous topic… can you also provide a ‘history lesson’, please? For instance, I read this just now and don’t know if you’ve done a Soft Reset / restore. It’s a very normal diagnostic step, you’ve been here 4 years, it’s very likely you’ve done all the steps. However, a little history of what you’ve done will stop you from getting repeats (I hope) :smiley:
Additionally, you can always re-migrate. That’s almost a factory reset in one click. :slight_smile:

Yes - I have done a soft reset … and restore … at least twice …. that is why i am thinking next of remigrating after a factory reset of that c8 pro … AFTER I prove out the c7 can deal with the new Z-wave devices (shades and lock) over the next week or so.

It was during this time that the c8 pro went south … mostly on just z-wave to start with … but wed morning it was just totally sluggish and very unresponsive … logs did not really show anything of substance.

I won’t try the c8 pro again until next week. Right now the little c7 is just humming away and everything is as snappy as heck.

I sure am glad I keep a change management log. All changes made to my home automation systems get documented. I could easily go back to may and update my c7 which got harvested back then so it has all the updates I made to the C8 pro.

One other difference - I was on beta program with the c7 - and then also added the c8 to that. I had problems with the http stuff on the betas after .159 so i fell back to 2.5.0.159 on the c8. I did not then update any beta software since …

On the c7 I am at 2.5.0.146 and will stay there for now …

History - I'm dealing with intermittent Z-Wave performance problems on a Hubitat C8 Pro that was migrated from a C7 using Hub Protect.

Symptoms included periods where many Z-Wave devices became slow or unresponsive, although Zigbee, LAN, and Home Assistant integration continued working normally.

I removed a ghost nodes from a failed z-wave inclusion of a new lock which is about when all this started.

I cleaned up stale devices, re-paired an Office SmartWings z-wave shade that had become very responsive after re-inclusion, and overnight Z-Wave logs generally looked healthy with no obvious retry storms or mesh collapse.

I did a couple soft resets and a couple restores to earlier states.

Wed morning - not only was z-wave performance bad, ANYTHING that called on the hub was extremely slow and unresponsive.

To isolate the cause, I've shut down the C8 Pro and redeployed my C7 with essentially the same database and devices. The C7 is currently running normally and is very responsive.

My plan is to run the C7 for several days, then add only a Yale YRD430 lock, and if it remains stable, perform a factory reset and fresh Hub Protect migration to the C8 Pro. If the freshly reset C8 Pro develops the same problems while the C7 remains stable, I'll consider that strong evidence of a C8 Pro hardware issue.

I have 5 Hubitat Hubs running all the time. Four C-8Pro that run my home (Upstairs, Downstairs, Front, and Internet Facing,) plus a C-8 as Development.

I chose this long ago, when my Hubs were all C-4’s, then C-5’s. Limiting the number of Z-Devices per hub was my intent. I have a self imposed limit of 65 devices per hub and it might be a factor in why I don’t see as many issues as the general community. It’s certainly NOT because I am in an interference free zone. :smiley: I’ve got 6 computers running in this room, Zigbee, BlueTooth, 3-USB sticks plugged into this computer, joined one per hub, in case I need to do Ghost Elimination… I haven’t needed that since the C-7, if I’m remembering right. If I had “RF Goggles” this room would be intense in RF radiation. LOL

Well - you would be a good role model on this - I have 67 z-wave devices NOW - 72 when/if i add those shades and z-wave lock. It wouldn’t be the worst idea in the world for me to have 2 hubs running …. will think on this.

my ‘fullest’ Hub is the only hub where I have both ZWave and Zigbee enabled. The other two have ZWave only and the internet facing hub has no radio enabled. (Although I think I have Matter enabled on all… not that I have many Matter devices, and those I have are Matter over WiFi.)

Opening Devices, it says I have 84 devices. 32 ZWave and 12 Zigbee (44 total Z-Radio devices) so the rest are virtual devices and I imagine no limit at all on those.

You may want to go back and read the release notes for Beta and consider trying again. There is something relevant to your situation(s) in the most recent release notes.

Oh yeah - I have Home Assistant tied to my hub - all the http problems they mentioned I had … I was on beta .112??? or so, and fell back to 2.5.0.159 which was stable for me ….

My performance problems started later …. but when i remigrate my c7 to c8pro fresh - it will have the same firmware as my c7 and I will stick there for awhile …

Seems like those were (mostly? all?) fixed in recent updates. However, you should read 2.5.1.129 notes specifically for your C8 Pro.

Many people, including me, have had experiences with a single bad Zwave device taking down the Zwave mesh and causing the hub to slow down or become unresponsive.

I would remove that device for a period of time and see if it improves things. Especially if dev:267 is a mains powered device that repeats.

And 2.5.1.130 beta has an additional Z-Wave update. Worth another look.

This seems like good advice, and easy to test w/out having to go back to the beta if you don't want to do that yet.

Yes - i DID have a bad z-wave device - but it wasn’t 267 … this one is a z-wave plug I had unplugged - it is now ok and plugged back in ….

I took the c8 pro back out last night, bumped it to the latest firmware - so far it has been working well … crossing fingers ….

Aren’t those z-wave updates all z-wave JS though? I AM now running .130 on std z-wave and it has been running very well for several hours - fingers crossed.

I upgraded the firmware to .130 beta, and then I just did one more soft reset and put a backup on there from 6-30 before all my problems. So far the hub seem to be happy … but don’t close this out yet guys …

Yes, I would strongly recommend running ZWaveJS at this point of time, especially because you have access to the updates because you're enrolled as a beta tester.

I’m gonna let this run a day or so … make a good backup again - then try a migration to Z-wave JS. Thx. I tried this awhile back DURING my problems tho … now without ?? problems - worth doing / trying again thx for the guidance.

Well - looks like all good. What I ended up doing …

  • I ran my old c7 hub for awhile with a backup i did right before migration.
  • I redid changes / updates I had made to my c8 which I document.
  • I WAS going to do a factory reset on my C8 and redo the migration but instead…
  • I removed a ghost z-wave from a failed inclusion (door lock)
  • I removed a chatty z-wave plug in module
  • I removed a chatty zigbee module.
  • I updated my C8 to the latest Beta firmware.
  • I did a soft reset and reloaded the backup from June 30 right before I was having my initial problems.
  • I updated the c8 hub with stuff I did since june 30 after ascertaining it seemed stable.
  • I ran logging on the system, z-wave, and zigbee overnight. Logs were perfect.
    Fed logs into AI - it gave my system an A-
  • Running VERY smoothly all day today.

Gonna call this resolved.

I just ordered a 2nd hub … I really like the idea of splitting the load and I have a large house …

My 4th hub is (as I mentioned) the internet facing hub, but it also where my Dashboards live. So I ‘mirror’ all the devices I need on a dashboard to that hub, from all the other hubs. And since the ‘mirrored’ devices are just virtual devices, they take no CPU power to manage. All of the Z-radios are disabled on that internet facing hub.

So… despite having, in my case, 212 devices, it’s got the lowest CPU % :smiley:

Therefore, I will go ahead and suggest you can do something similar since you have a ‘spare’ C-7 now :smiley: Let it be the way to connect to Alexa/Google Home, Weather, etc. via the Internet and also be your Dashboard that is available via Hubitat’s cloud. So when you are hanging about in MacDonalds and need to know if the spa is ready, it’s right there. :smiley: :smiley: :smiley: