#ZWA-2 1.2.0 Timeout Waiting for Response

1 messages · Page 1 of 1 (latest)

lofty jacinth
#

Seems like I did the 1.2.0 firmware and now nearly all of my switches a combination of GE Enlightened Gen 5 and Zooz Gen 7 do not work. All devices fail to interview.

I tried to shut down home assistant; I rebooted the ZWA multiple times. I tried to use the web toolbox. When I clicked on recover it said device is functioning fine. When I tried to flash to latest firmware it errors out.

I have attached a log file. I plugged in my old Zooz stick and it took right off.

I took an NVM backup of the Zooz stick and I plugged in the ZWA did a factory reset and restored the NVM file. Same results is what you see in the logs.

coarse lava
little nimbus
#

I think that's a different issue. OP says the recovery claims the ZWA-2 is working.

#

The communication pattern is really strange. the numbers mostly look good, commands even get acknowledged but the responses are not received most of the time. Can you try power cycling one of the problematic devices to see if that helps?

lofty jacinth
#

Tried that and didn’t work. The issue seems to be devices that are setup with S2. It’s almost like the S2 keys changed.

└─[Security2CCMessageEncapsulation] [INVALID]
error: Security2CC_NoSPAN
2026-02-27 18:29:54.827 CNTRLR » [Node 163] No SPAN is established yet, cannot decode command. Requesting a non

If I feed the log to Gemini, Claude or ChatGPT they all come to the conclusion that there is an issue with the secure layer

From Gemini

“The core issue is happening at the Security 2 (S2) protocol layer. The controller and your devices are failing to establish a secure session (called a SPAN).
• The controller explicitly flags this, logging: No SPAN is established yet, cannot decode command. span_3Requesting a nonce....
• Because the secure handshake is failing, the controller cannot decrypt incoming messages. It actively rejects them, logging error: Security2CC_NoSPAN and Dropping message with invalid payload.”

So I’m at a loss as to what to try next.

#

The S2 keys didn’t change. They are the same S2 setup. I can plug in the Zooz 800 series chip does on SDk 7.2x and everything works no issues.

Did SDK 8.0.0 change how security keys work? Does it regenerate on first boot or something?

lofty jacinth
#

https://www.silabs.com/documents/public/release-notes/SRN14930-7.20.0.0.pdf

Section 2 here.

“Note that due to the introduction of Secure Key Storage on the 800 series, having externally supplied key pairs is no longer
supported. To ensure that security is not compromised, keys are generated internally on first boot and kept only in secure storage.”

Which all of this sounds like to me… I need to redo my entire network.

azure shard
#

You linked to an SDK version that's older than what was ever installed on the ZWA-2, and doesn't apply to controllers.

lofty jacinth
#

Yes it wasn’t about the SDK version. It shows that the 800 series chip use Secure Key Storage. All manuals since then state the same thing.

little nimbus
#

Keys are configured in home assistant and not stored in the ZWA-2. Your issue is consistent with wrong keys for what it's worth

#

But no firmware update is going to cause that

lofty jacinth
#

So how the same security keys work on one stick but not a second stick? Both 800 series chips. Zooz 800 vs ZWA-2. System was initially on the Zooz 800 migrated to the ZWA-2. ZWA has worked since September. But no longer works now.

little nimbus
#

Doesn't really make sense. Unless they somehow got deleted.

#

The controller doesn't know anything about encryption. All it does is send data back and forth.

lofty jacinth
#

I can plug in a device. Switch it in the ZWave menu. Restart JS UI. New device takes over. Keys never move or change.

lofty jacinth
#

I tried to use command prompt. Or zwave PC Controller OTW updates.

Each time we get a 0x44 error.

HaOS terminal to do the update. Windows terminal to update. I also did try and delete clear the NVM again.

I was thinking it was a failed firmware update but I can’t even do a new firmware update to try and correct it. I have attached the logs from the different attempts maybe it will help someone.

little nimbus
lofty jacinth
#

Yeah. I will have that tonight.

lofty jacinth
lofty jacinth
#

The ZWA isnt dead. It can control items that are setup as No Security. See these logs for Node 173. Was a ceiling fan and the speed was adjusting despite it saying it timed out.

little nimbus
#

I don't see any log from the Zooz stick where you are actually controlling the same devices. Only one with a firmware update going on.

I noticed two things in the ZWA-2 logs:

  1. The powerlevel settings are not what they should be (too high for mesh, too low for LR). You should be able to correct this by temporarily switching the region to something else and back.
    Open the Web UI, then:
#

Afterwards I recommend rebuilding the routes because this is likely going to change some routes.

#
  1. None of the problematic devices are actually responding to commands. This might be because of the incorrect powerlevel settings, but for some reasons the acknowledgements are received just fine. Rebuilding routes might fix this too.
    In the example where you said the devices are reacting despite timing out, the command was sent with Supervision, which requires the device to respond with an update whether the command was executed. This didn't happen in your log, hence the timeout.
lofty jacinth
#

Here are logs of a routeRebuild seems like the ZWA2 doesn’t want to rebuild routes. Just hangs. I don’t see it climbing in controllers stats for rx or tx.

I did the same with the Zooz and thst worked.

I did a new log with the Zooz (zwave_log) node 113, 117, 119, 121. This is with Zooz.

The node region for the ZWA was not set at all. So o set that to UsA long range and sent. It took.

little nimbus
#
2026-03-03 12:20:04.429 DRIVER « [REQ] [RequestNodeNeighborUpdate]
                                   callback id:   143
                                   update status: UpdateFailed

Looks like a symptom of the same issue - the ZWA-2 likely isn't receiving the responses after the route rebuild.

#

Did you try to do the route rebuild after the region setting again?

#

Because I now see commands being received:

2026-03-03 12:36:47.330 DRIVER « [Node 163] [REQ] [BridgeApplicationCommand]
                                 │ RSSI: -79 dBm
                                 └─[Security2CCMessageEncapsulation]
                                   │ sequence number: 142
                                   │ security class:  S2_Authenticated
                                   └─[MultiChannelCCCommandEncapsulation]
                                     │ source:      8
                                     │ destination: 0
                                     └─[MeterCCReport]
                                         meter type: Electric
                                         scale:      W
                                         rate type:  Produced
                                         value:      0
                                         time delta: 0 seconds
little nimbus
#

Also the reason that I told you to change it to a different region first was to make sure the automatic powerlevel adjustment is done.

lofty jacinth
#

I changed to Europe Long Range. It took that command.

I changed back to USA Long Range. It took that command.

Attempting to rebuild routes.

Node 08 rebuilding routes failed: could not update the neighbors list after 5 attempts.

Now it is on Node 10. And those are also failing. It has been 9 minutes an I am had 5 failed attempts on Node 8 and 4 failed attempts on Node 10.

At this point I am just going to try and ask the company I bought the device from if they will warranty it. Becasue something is clearly broke with this.

little nimbus
#

Can you make an NVM backup and send me that file so I can take a look if anything stands out?

lofty jacinth
little nimbus
#

Nothing that really jumps out. Were you also unable to control LR devices or just mesh?