search

magnific0 / wondershaper

Command-line utility for limiting an adapter's bandwidth
GNU General Public License v2.0
1.73k stars 264 forks source link

Ubuntu Mate 16.04 BCM2709 / BCM2836 continuous "hw csum failure" exceptions #19

Closed madaello closed 4 years ago

madaello commented 6 years ago

syslog.example.log Version: Distributor ID: Ubuntu Description: Ubuntu 16.04.3 LTS Release: 16.04 Codename: xenial

Network: 2: enxb827eba318dc: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000 link/ether b8:27:eb:a3:18:dc brd ff:ff:ff:ff:ff:ff inet 192.168.0.131/24 brd 192.168.0.255 scope global dynamic enxb827eba318dc valid_lft 84219sec preferred_lft 84219sec inet6 fe80::7975:a397:be6c:b6a/64 scope link valid_lft forever preferred_lft forever

adapter used: enxb827eba318dc

Around 300 syslog lines every 5 seconds. I don't think that is normal behaviour, do you have an explanation/solution?

magnific0 commented 6 years ago

That is indeed very aggressive. I am surprised your network stays up as it seems to continuously crash and restart.

I don't think this problem will can be solved easily as the script just add simple tc rules, which shouldn't cause this behaviour. There is definitely a bug in the kernel / broadcom driver here. My only suggestion is to update both to the latest version and if still persistent file a bug report with the either Ubuntu Mate or rpi linux.

magnific0 commented 6 years ago

Could you please report the output of uname -a here, so we have an idea of the kernel you're one. Thanks!

madaello commented 6 years ago

about uname -a Linux umate-pi 4.4.38-v7+ #938 SMP Thu Dec 15 15:22:21 GMT 2016 armv7l armv7l armv7l GNU/Linux

I ran the same script this night (basically a curl script to upload files to dropbox) without limit the bandwidth and I didn't see any kernel error in syslog. However no worries, I can stay without limit the bandwidth. I just advised you of this strange behaviour in case you have a simple solution, because at the end wondershaper did the correct job but with tons of kernel lines in syslog.

magnific0 commented 4 years ago

Sorry for the late reply. I appreciate your feedback. However, I'm closing this issue as it is quite stale. If tc is (still) spamming the syslog this should really be taken up there, as the resources on this project are very limited. Thank!