scudderfish
Bit poster
llipd appears to be opening 2 UDP sockets per second continuously until my Mac keels over with "Aug 2 17:53:46 gedge kernel[0]: file: table is full" when it has managed to grab all 12000 file handles. I've got 4560 installed, and this happens even after a hard reboot (I have to hit the power switch to recover) and I've not run any VMs
It's currently at 1018 open handles and counting. I fingered llipd as the culprit by running
gedge:~ root# while (true) do date >> lsof.log; lsof | cut -d ' ' -f 1 | sort | uniq -c | sort >> lsof.log; sleep 60; done
and then
gave me lots of
llipd 140 root 1057u IPv4 0x4ae9930 0t0 UDP *:*
llipd 140 root 1058u IPv4 0x4ae9860 0t0 UDP *:*
llipd 140 root 1059u IPv4 0x4ae9790 0t0 UDP *:*
llipd 140 root 1060u IPv4 0x4ae96c0 0t0 UDP *:*
llipd 140 root 1061u IPv4 0x4ae95f0 0t0 UDP *:*
llipd 140 root 1062u IPv4 0x4ae9520 0t0 UDP *:*
llipd 140 root 1063u IPv4 0x4ae9450 0t0 UDP *:*
llipd 140 root 1064u IPv4 0x4ae9380 0t0 UDP *:*
llipd 140 root 1065u IPv4 0x4ae92b0 0t0 UDP *:*
gedge:~ root# lsof | grep llipd
I'm going to kill -9 140 now and see if my Mac stays up for the rest of the day
It's currently at 1018 open handles and counting. I fingered llipd as the culprit by running
gedge:~ root# while (true) do date >> lsof.log; lsof | cut -d ' ' -f 1 | sort | uniq -c | sort >> lsof.log; sleep 60; done
and then
gave me lots of
llipd 140 root 1057u IPv4 0x4ae9930 0t0 UDP *:*
llipd 140 root 1058u IPv4 0x4ae9860 0t0 UDP *:*
llipd 140 root 1059u IPv4 0x4ae9790 0t0 UDP *:*
llipd 140 root 1060u IPv4 0x4ae96c0 0t0 UDP *:*
llipd 140 root 1061u IPv4 0x4ae95f0 0t0 UDP *:*
llipd 140 root 1062u IPv4 0x4ae9520 0t0 UDP *:*
llipd 140 root 1063u IPv4 0x4ae9450 0t0 UDP *:*
llipd 140 root 1064u IPv4 0x4ae9380 0t0 UDP *:*
llipd 140 root 1065u IPv4 0x4ae92b0 0t0 UDP *:*
gedge:~ root# lsof | grep llipd
I'm going to kill -9 140 now and see if my Mac stays up for the rest of the day