Hi,
today during two hours kamailio made 2G of a syslog messages. This message was repeated for 10 000 000 times.
/sbin/kamailio[24272]: ERROR: <core> [tcp_main.c:3959]: WARNING: handle_new_connect: error while accepting connection(24): Too many open files
It was preceded by following error.
/sbin/kamailio[24265]: ERROR: <core> [tcp_read.c:269]: error reading: Connection reset by peer (104) /sbin/kamailio[24265]: ERROR: <core> [tcp_read.c:882]: ERROR: tcp_read_req: error reading
Finally it stopped without any given reason.
I'm using kamailio 3.1.0 (i386/linux) 1e204f, callcontrol 2.0.11 and clustered mysql.
Was this some kind of attack, bug, error, misconfiguration or what?
Thank for answer,
Efelin
On Friday 11 February 2011, Efelin Novak wrote:
today during two hours kamailio made 2G of a syslog messages. This message was repeated for 10 000 000 times.
/sbin/kamailio[24272]: ERROR: <core> [tcp_main.c:3959]: WARNING: handle_new_connect: error while accepting connection(24): Too many open files
It was preceded by following error.
/sbin/kamailio[24265]: ERROR: <core> [tcp_read.c:269]: error reading: Connection reset by peer (104) /sbin/kamailio[24265]: ERROR: <core> [tcp_read.c:882]: ERROR: tcp_read_req: error reading
Finally it stopped without any given reason.
I'm using kamailio 3.1.0 (i386/linux) 1e204f, callcontrol 2.0.11 and clustered mysql.
Was this some kind of attack, bug, error, misconfiguration or what?
Hi Efelin,
could be an attack, but maybe its just an small misconfiguration of your server? Have you tried to increase the "ulimit -n" and "/proc/sys/kernel/file- max" settings? Then your box should be able to open more file handles.
Cheers,
Henning