A number of problems (including bugs?) break firmware update

Configuration Z-Wave
1

I have a growing network with 17 Z-Wave nodes, mostly wired (only one battery sensor) and almost all of them from Aeotec (including Z‐Stick Gen5 adapter). Most of the nodes are included through SmartStart.

I’m having some communication problems that become most apparent when doing firmware update through Z-Wave JS UI (I also tried the regular UI, same result).

Looking at the logs (https://pastebin.com/raw/Ztcvd816), there’s a few problems:

  1. When the update goes, it goes smoothly (sending fragments every 100-500ms), but if it trips (and it trips often), it gets stuck for couple or even dozens of seconds
  2. If any other sensor updates its values, it trips with “CNTRLR Failed to execute controller command after 1/3 attempts. Scheduling next try in 100 ms.”
  3. Occasional “DRIVER Dropping message with invalid payload”
  4. Sometimes “received S2 nonce without an active transaction, not sure what to do with it” on otherwise healthy nodes
  5. “Firmware update timed out” after 34 seconds from the last sent firmware fragment (but having “Dropping message with invalid payload” in the meantime, which could be the fw request?)
  6. “Received Firmware Update Get, but no firmware update is in progress. Forcing the node to abort…” 5 seconds after firmware update timeout (looks like the node still wanted another fragment)

My wild guess is that problem 3 is the root cause, but extending the timeout (problem 5) would at least unblock firmware updates.

Any ideas?

2

I got the problem fixed by replacing Aeotec Z-Stick Gen5 adapter with Zooz 800 Series LR.

I don’t know if it’s a matter of going from 500-series stick to 800-series (all of my nodes are 700-series), or changing the manufacturer, but now the network is noticeably more responsive and firmware updates never break. I suspect the former, because I still see problems 1-4, but either less often or they trip the process for shorter, because it never hits the timeout. I have an impression this gets worse as I grow my network.

Before swapping controller, I tried turning down all sensors (so it was only hardwired switches and dimmers) and turning down the two non-Aeotec nodes (by just flipping the breaker for the part of house they were wired to) - no difference.

EDIT: I just hit another failing node (one of the farthest, but absolutely not the farthest one). Seems like the new stick made everything better, but not resolved the problem entirely.

3

Having the same problem and I have the Zooz 800 LR USB