NWA50AX Pro - Random Reboots & "Internal hardware watchdog timeout" under load

Options
rosariodellib
rosariodellib Posts: 4 image  Freshman Member
First Comment Friend Collector

I am opening this thread to report a critical instability issue affecting two NWA50AX Pro units.Both devices experience random spontaneous reboots. The system log consistently reports the following error:Device is rebooted by internal hardware watchdog timeout!

Environment:

  • Devices: 2x NWA50AX Pro.
  • Management: Currently in Standalone Mode (I factory reset them and moved away from Nebula to rule out cloud config issues).
  • Power Source: Zyxel GS1005HP PoE+ Switch (Brand new, replaced specifically to rule out power issues).
  • Network Load: The Wi-Fi network handles constant video streaming from 8 IP Cameras.

Troubleshooting steps already performed (Issue persists):To isolate the root cause, I have already tried the following conservative configurations, but the Watchdog still triggers a reboot:

  1. Hardware: Replaced the PoE Switch and all Ethernet cables.
  2. Radio Configuration: Reduced 5GHz Channel Width from 160MHz to 80MHz (and even 40MHz) to reduce CPU load.
  3. CPU Offloading: Switched from "Multicast to Unicast" to Fixed Multicast Rate (configured at 11Mbps for 2.4G and 24Mbps for 5G) to relieve the CPU from packet conversion tasks.
  4. Features Disabled: Smart Mesh is OFF, DCS is OFF, 802.11r is OFF.

Conclusion:Since the issue happens on two different units and persists even after replacing the power source and minimizing CPU load via configuration, this appears to be a Firmware/Driver instability (likely a memory leak or driver freeze handling constant UDP/Multicast streams).

Request for Support:

  1. Is there a specific stable or Beta Firmware available that addresses this Watchdog Timeout / CPU Freeze issue?
  2. I have the diag-info files ready from the Standalone GUI. Should I upload them here, or do you require me to onboard the APs back to Nebula Cloud for remote inspection?

Looking forward to a solution as the network is currently unstable.

Thanks.