If I do umount /dev/io-net/en1 I get the following entry in slog that comes up every 5 sec (period_ticks is set to 10 ) lr_tk_pkt() bad l4_index: 1 Is this something I should worry about? - Mario
No. The qnet resolver is trying to transmit on an L4 which no longer exists, because you unmounted it, and is whining about it. If it's swamping your slogger, you can run qnet with the "no_slog=1" option, which still allows qnet's internal event ringbuffer to work (see cat /proc/qnetstats).