When can we expect new firmware for NWA50/55/90AX(E)?

Hi Kay,

I've been through so many attempts to resolve the key handshake failure problem, but the only successful fix for me is a simultaneous reboot of all APs on the network. This seems to clear the "shared" client information in the APs and lets the affected clients back on the network.

I also tried to create a dedicated 2.4Ghz IoT SSID, but it caused so much instability in the APs that I had to remove it within about an hour of creating it. Some APs rebooted and refused to broadcast any SSIDs until the new IoT SSID was removed from Nebula. I will attempt it again, when I can get a suitable "maintenance window" without annoying the rest of my family!

I keep a regular eye on the https://community.zyxel.com/en/categories/wlan-news-and-release page already and the number of firmware releases this year (so far) is much lower than last year (for example).

When you say "Q4 this year", I will assume you mean "Oct-Dec" (we in the UK have several different yearly calendars so it can be ambiguous!).

Thanks,

Lewis.

Hi Kay,

I have now created an "IoT" SSID (after negotiating a Wi-Fi maintenance downtime at home), which is 2.4Ghz only with no roaming assistance options enabled.

Since doing this, one of my APs is spamming my SYSLOG server (about 12-15 messages a minute) with messages like this

kernel: [xxxx.xxxx] illegal BA buffer size = 0

kernel: [xxxx.xxxx] MakeCountryIe: reg_class is NULL !!!

In fact, total Syslog usage appears to be up since creating the SSID

The massive spike around 18:00 is when the SSID was created, and the small spike around 2am is when my scheduled reboot of all APs occurs.

Annoyingly, the APs all give the hostname as their model name (rather than the name that I assign each unit), so all the NWA90AX logs are merged in to one in my log viewer (just using Synology's built in log centre, so nothing fancy).
This is different to the Zyxel GS1920 network switches I have, which give their hostname as the name I have assigned them, so I can separate them in the log centre.

Also, the APs don't seem to "confirm" to the BSD SYSLOG format, as the data is not matching the fields correctly (but could be Synology bug)

GS1920

NWA90AX

I haven't tried using IETF format yet (would require creating a new listener on a different port in the Synology), but then why would Zyxel use one SYSLOG format for network switches and a different format for APs?

My Roomba tried to reconnect to the IoT network after the APs were rebooted overnight, but kept getting kicked off due to

Station: 80:a5:89:xx:xx:xx disconnected by Configuration Changed on Channel: 6, SSID: xxxxx-IoT, 2.4GHz, Signal: -40dBm, Download/Upload: 4769/3554 Bytes, reason 4, Interface: wlan-1-4

So although the APs start advertising the SSID, it seems to keep making configuration changes for a short while after which kicks everything off: surely the AP should only start advertising the SSID when it is ready to do so after a reboot? It knows what the configuration is, so shouldn't have to keep altering the SSID configuration. It should be "set up SSID configuration then advertise", not "advertise SSID then set up the configuration".

Or does an SSID configuration change get triggered when a new AP joins the network? During my scheduled reboot, they will all re-join at staggered times and maybe cause this? However, the IoT SSID is configured with no roaming assistance options, so not sure why it would be affected by the network gaining new APs.

Anyway, I will need to reboot the Roomba to get it back on the network (it gets grumpy) and I will keep monitoring the IoT network performance.

Lewis.

Hi @Zyxel_Kay,

And suddenly this evening, my Amazon Echo has been kicked off the Wi-Fi with the dreaded "key handshake fail" message.

It hunts around between all the APs, and even switches Bands & SSIDs (I told it to move to the IOT SSID but Amazon remembers the AP SSID, even though I told Amazon to forget my Wi-Fi networks … apparently I have to deregister/re-register the device to clear it down. But that is a completely different problem).

All APs were rebooted at 2am last night, and I bet if I reboot them all again now the Amazon Echo will join the network again without any changes to it.

Is there any other information I can gather to help resolve this painful problem?

Thanks,
Lewis.

Hi @Zyxel_Kay,

I have enabled Zyxel Support access, and PM'd you the details of my org/site.

Thanks,

Lewis.