3

I DDoS'ed my pi as part of a stress test using 2 laptops, fully saturating the 150 Mbps LAN interface on it. Less than an hour later, I found out that the pi doesn't respond to the ping flood and TCP flood anymore. I double checked and saw that the interface doesn't work until I restarted the networking service. When I checked dmesg (before restarting networking), I found a lot of these:

smsc95xx 1-1.1:1.0 eth0: kevent 0 may have been dropped

Anyone know what this is? Is this normal? I've Googled around and it seems that people mostly encounter a kevent 2 instead of this kevent 0. That's why I couldn't find any info.

I didn't touch any of the networking stuff except when I set a static IP address. All is well when network activity isn't as flooded as this.

Stable overclock: 1000MHz ARM, 500MHz core, 450MHz SDRAM, 4 overvolt, Heatsink and fan on SOC and LAN chips

  • I get this a lot now when trying to pull docker images running ubuntu server 16 on my Pi3. What did you do to fix it? – Kendall ARNEAUD Jan 19 '18 at 12:47
  • @Kendall I didn't do anything. It happens when there's way too much traffic. On the link in the answer, at line 450, is where the error comes from. The conditions (if (!schedule_work and if (net_ratelimit())) tell why this happens. The best action, I think, is to simply free up some resources to let the kernel handle network traffic. – PNDA Jan 20 '18 at 6:12
2

It seems the kernel couldn't handle it, based on the source code.

usbnet.c on Github

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.