High CPU usage = broadcast storm?

[badbob001]

I have a few AC68U (AP mode) running 386.14_2. Recently, my LAN performance is rather poor and I noticed all my AP have 100% Core 1 usage. When I unplug my hdhomerun tuner from an unmanaged network switch, the cpu usage instantly drops.

What does this mean? I'm guessing the hdhomerun was creating network traffic that my AP need a lot of CPU to handle. Maybe the hdhomerun was faulty and caused a broadcast storm? What else would cause high cpu usage on the AP? I just factory reset them with minimal changes. I don't use aimesh.

 

[bennor]

Connect to the router via SSH and run top to see what is consuming the router's resources.

 

[badbob001]

Used wireshark and saw that my firetv cube and hdhomerun were blasting each other with MDNS (multicast dns) messages. It looks like the firetv sent the first query. No idea why only these two devices are behaving this way to each other suddenly.

I didn't see a way to disable mdns on either device. I unplugged the network cable from the firetv and replugged it back in after awhile, issue came back. Power-cycled the firetv. Okay so far.

 

[Pierre Nakashian]

I had the same 1 high cpu core problem when I woke up this morning, the high cpu also caused my wifi pretty much to be useless. I figured out this evening after returning from work it was hdhomerun, after reading your post i also narrowed it down to 1 FireTV Cube, unplugging didn't help, i'm factory resetting the FireTV cube, see it that helps.

update:
factory reset worked, interesting I couldn't install any apps from the app store because I have Adguard running, I had to temporary put bypass rules install the apps and put the dns global rules back.

 

[badbob001]

Silicondust support told me to try the January 2025 hdhr beta firmware, which has many changes related to mdns. I updated my hdhr devices and it resolves the issue. Can't really say if that means it's an issue with the hdhr as the last stable firmware release was from 2023. So something must have recently changed on the cube. Probably something I don't care about like alexa+ or matter.

Perhaps this issue mostly affects users here since who else would still be using wifi AP's from more than 10 year ago that don't have the CPU performance to handle a multicast flood. Then again, maybe some of us have switches that can curb these issues.

Was the issue with a cube 3? I have mine connected to ethernet via usb-to-ethernet dongle. I wonder if it were on wifi, would there be better management of the mdns flood. The cube would certainly knock itself off the wifi network...

 

[Pierre Nakashian]

happened again this morning, same FireTV cube I factory reset and did all the upgrades. I put the FireTV cube behind an ASUS rtac68u access point and ran. "service stop_mdns" also put that in startup script services-start. crossing my fingers. I'll also try the fw upgrade on hdhr. I just submitted a ticket to Silicondust support to get that same Firmware.

 

[Pierre Nakashian]

Silicondust support told me to try the January 2025 hdhr beta firmware, which has many changes related to mdns. I updated my hdhr devices and it resolves the issue. Can't really say if that means it's an issue with the hdhr as the last stable firmware release was from 2023. So something must have recently changed on the cube. Probably something I don't care about like alexa+ or matter.

Perhaps this issue mostly affects users here since who else would still be using wifi AP's from more than 10 year ago that don't have the CPU performance to handle a multicast flood. Then again, maybe some of us have switches that can curb these issues.

Was the issue with a cube 3? I have mine connected to ethernet via usb-to-ethernet dongle. I wonder if it were on wifi, would there be better management of the mdns flood. The cube would certainly knock itself off the wifi network...

I applied the HDHR firmware update and removed the "service stop_mdns" network is back to normal. I have a great switch that can curb some multicast traffic, it works great on a smartthings hub, but even the switch couldn't curb the FireTV cube and HDHR combined multicast traffic. I have more than 1 FireTV Cube, only 1 of them was causing the issue, factory reset and software updates didn't fix it. I'm going to guess this FireTV cube only has received a new amazon software update, the timing of both of our issues is not a coincident. Amazon probably staggers software updates to FireTV cubes so the downloads don't overwhelm their servers at the same time.