Hi, I'm running an entry-node on a virtual-server. I have usually a throughput of 200-400 GB each day and it runs fine so far. My problem is that my tor-logs are spammed with these messages: .. Aug 07 09:39:57.440 [warn] Error from libevent setting read event state for 1023 to watched: No such file or directory Aug 07 09:45:25.996 [warn] Error from libevent setting read event state for 1023 to watched: No such file or directory Aug 07 09:50:25.213 [warn] Error from libevent setting read event state for 1023 to watched: No such file or directory Aug 07 09:55:26.592 [warn] Error from libevent setting read event state for 1023 to watched: No such file or directory Aug 07 10:08:33.343 [warn] Error from libevent setting read event state for 1023 to watched: No such file or directory Aug 07 10:16:00.721 [warn] Error from libevent setting read event state for 1023 to watched: No such file or directory Aug 07 10:21:27.403 [warn] Error from libevent setting read event state for 1023 to watched: No such file or directory Aug 07 10:26:35.141 [warn] Error from libevent setting read event state for 1023 to watched: No such file or directory Aug 07 10:31:35.452 [warn] Error from libevent setting read event state for 1023 to watched: No such file or directory Aug 07 10:38:19.442 [warn] Error from libevent setting read event state for 1023 to watched: No such file or directory Aug 07 10:44:33.383 [warn] Failing because we have 994 connections already. Please raise your ulimit -n. Aug 07 10:46:05.720 [warn] Error from libevent setting read event state for 1023 to watched: No such file or directory Aug 07 10:52:35.242 [warn] Error from libevent setting read event state for 1023 to watched: No such file or directory Aug 07 13:05:43.256 [warn] Error from libevent setting read event state for 1023 to watched: No such file or directory Aug 07 13:12:24.525 [warn] Error from libevent setting read event state for 1023 to watched: No such file or directory Aug 07 13:57:00.948 [warn] Error from libevent setting read event state for 1023 to watched: No such file or directory .. Here are some informations about my server: $ ulimit -H -n 1024 $ cat /proc/user_beancounters Version: 2.5 uid resource held maxheld barrier limit failcnt 33209: kmemsize 3380633 3415768 79691776 88080384 0 lockedpages 0 0 3584 3584 0 privvmpages 21901 22100 551680 559872 0 shmpages 640 640 131072 131072 0 dummy 0 0 0 0 0 numproc 10 10 1536 1536 0 physpages 16377 16437 0 9223372036854775807 0 vmguarpages 0 0 142080 9223372036854775807 0 oomguarpages 16377 16437 142080 9223372036854775807 0 numtcpsock 994 996 1792 1792 0 numflock 2 2 1000 1100 0 numpty 1 1 64 64 0 numsiginfo 0 0 1024 1024 0 tcpsndbuf 3744392 4346800 16777216 25165824 0 tcprcvbuf 2096520 2165928 16777216 25165824 151 othersockbuf 14064 15384 8388608 15728640 0 dgramrcvbuf 0 8488 8388608 8388608 0 numothersock 11 13 1792 1792 0 dcachesize 1092906 1108890 16777216 17301504 0 numfile 1538 1538 16384 16384 0 numiptent 14 14 256 256 0 On the server runs nothing except tor and a ssh-server so all rescources are used by tor. I think the problem is with the tcprcvbuf because it has a failcount of 151. But I don't know what to do now. How do I stop these messages/problems with my node? What limits should I set in my torrc? Raising tcprcvbuf is not possible for me. Shall I reduce the ConnLimit or the ConstrainedSockSize? What would you advice me? cu
Attachment:
signature.asc
Description: PGP signature