You are not logged in.
I was watching the Kernel Logs/Kernel Ring Buffer using
sudo dmesg -wto debug an unrelated issue and I happened to notice that majority of the log messages are the following:
[75200.551003] IPv6: MLD: clamping QRV from 1 to 2!
[75206.553054] IPv6: MLD: clamping QRV from 1 to 2!
[75212.557678] IPv6: MLD: clamping QRV from 1 to 2!
[75218.559370] IPv6: MLD: clamping QRV from 1 to 2!
[75224.563031] IPv6: MLD: clamping QRV from 1 to 2!
[75230.565733] IPv6: MLD: clamping QRV from 1 to 2!
[75236.569101] IPv6: MLD: clamping QRV from 1 to 2!
[75242.573379] IPv6: MLD: clamping QRV from 1 to 2!
[75248.576252] IPv6: MLD: clamping QRV from 1 to 2!
[75254.579186] IPv6: MLD: clamping QRV from 1 to 2!Now, there are solutions to fix this online which just involve running a magic command to fix it, but I need to understand what is the source of this issue (if it is an issue), what does this line mean and what does a fix do. I couldn't find details about this on IPv6 page of the Arch Wiki and most online documentation for this is either too high level or too theoretical. Please help me understand what is going on here.
Offline
According to an explanation I have found while searching for this error message this "robustness value" for spotty connections is set to 2 (default) and regards a value of "itself minus 1" packet losses as robust. It must not be zero and should not be 1 (which is the solution you found).
As I understand it, with expected spotty connections you can set this value even higher (say 10) to regard 9 packet losses as acceptable.
The problem is, that some process is lowering your value to 1 and the kernel is correcting it. It would be interesting what causes the decrease.
Offline
How can I get further details about what it is, that is setting this value to 1?
Offline
system journal, active (networking) services, "ip a"
Walking up the linear calltree of https://elixir.bootlin.com/linux/v7.1.4 … st.c#L1284 this has to come from https://elixir.bootlin.com/linux/v7.1.4 … onf.c#L377
With your frequency (every 6s) this heavily smells like conflicting network services reconfiguring the NICs but there might be overall issues w/ the IPv6 stack (if you don't have/need/can't use ipv6 you'd simply disable that altogether)
Offline