[OpenAFS] fileserver 1.2.11 problem
Thomas Mueller
thomas.mueller@hrz.tu-chemnitz.de
Tue, 12 Oct 2004 14:57:16 +0200 (MEST)
Hi all,
we have got a problem with some of our fileservers (OpenAFS-1.2.11,
LWP fileserver, RedHat Linux 7.3).
The fileserver process cycles and doesn't reply on requests. It even doesn't
respond on signals other than -KILL or -ABRT.
One day I got a core dump (from kill -ABRT) and the backtrace shows:
(gdb) where
#0 0x420e8412 in sendmsg () from /lib/i686/libc.so.6
#1 0x0808a1a3 in rxi_Sendmsg ()
#2 0x08092042 in osi_NetSend ()
#3 0x08092d8e in rxi_SendPacket ()
#4 0x0808e157 in rxi_SendList ()
#5 0x0808e368 in rxi_SendXmitList ()
#6 0x0808e6ea in rxi_Start ()
#7 0x080964f8 in rxevent_RaiseEvents ()
#8 0x08089cd8 in rxi_ListenerProc ()
#9 0x0808a014 in rx_ServerProc ()
#10 0x080982ac in Create_Process_Part2 ()
#11 0x080988e6 in savecontext ()
(gdb)
tcpdump -s 1500 -vv port 7000
looks like this:
6:07:42.315566 pleione.hrz.tu-chemnitz.de.afs > 235.233.107.65.12544: [udp sum
ok] rx data cid 059f3524 call# 52267 seq 1 ser 356513307 <client-init>,<req-ack
>,<last-pckt> (32) (DF) [ttl 1] (id 0, len 60)
06:07:42.315589 pleione.hrz.tu-chemnitz.de.afs > 235.233.107.65.12544: [udp sum
ok] rx data cid 059f3524 call# 52267 seq 1 ser 356513308 <client-init>,<req-ack
>,<last-pckt> (32) (DF) [ttl 1] (id 0, len 60)
06:07:42.315598 pleione.hrz.tu-chemnitz.de.afs > 235.233.107.65.12544: [udp sum
ok] rx data cid 059f3524 call# 52267 seq 1 ser 356513309 <client-init>,<req-ack
>,<last-pckt> (32) (DF) [ttl 1] (id 0, len 60)
06:07:42.315607 pleione.hrz.tu-chemnitz.de.afs > 235.233.107.65.12544: [udp sum
ok] rx data cid 059f3524 call# 52267 seq 1 ser 356513310 <client-init>,<req-ack
>,<last-pckt> (32) (DF) [ttl 1] (id 0, len 60)
06:07:42.315616 pleione.hrz.tu-chemnitz.de.afs > 235.233.107.65.12544: [udp sum
ok] rx data cid 059f3524 call# 52267 seq 1 ser 356513311 <client-init>,<req-ack
>,<last-pckt> (32) (DF) [ttl 1] (id 0, len 60)
...
Since we 've got the problem quite often (one time per day and server), I 've
installed the following rules for iptables:
:INPUT DROP [0:0]
:OUTPUT DROP [0:0]
...
-A INPUT -p udp -m udp --dport 7000 --sport 7001 -j ACCEPT
...
-A OUTPUT -p udp -m udp --sport 7000 --dport 7001 -j ACCEPT
...
other UDP traffic will be logged and I see the following entries:
Oct 12 07:57:26 pleione kernel: iptables REJECT:IN= OUT=eth0 SRC=134.109.132.140 DST=230.141.0.0 LEN=60 TOS=0x00 PREC=0x00 TTL=1 ID=0 DF PROTO=UDP SPT=7000 DPT=17552 LEN=40
Oct 12 07:57:42 pleione kernel: iptables REJECT:IN= OUT=eth0 SRC=134.109.132.140 DST=230.141.0.0 LEN=93 TOS=0x00 PREC=0x00 TTL=1 ID=0 DF PROTO=UDP SPT=7000 DPT=17552 LEN=73
...
and later until the next fileserver cycle:
Oct 12 11:42:04 pleione kernel: iptables REJECT:IN= OUT=eth0 SRC=134.109.132.140 DST=184.11.109.65 LEN=60 TOS=0x00 PREC=0x00 TTL=64 ID=0 DF PROTO=UDP SPT=7000 DPT=29191 LEN=40 Oct 12 11:42:20 pleione kernel: iptables REJECT:IN= OUT=eth0 SRC=134.109.132.140 DST=184.11.109.65 LEN=93 TOS=0x00 PREC=0x00 TTL=64 ID=0 DF PROTO=UDP SPT=7000 DPT=29191 LEN=73
Oct 12 12:10:33 pleione kernel: iptables REJECT:IN= OUT=eth0 SRC=134.109.132.140 DST=94.18.109.65 LEN=93 TOS=0x00 PREC=0x00 TTL=64 ID=0 DF PROTO=UDP SPT=7000 DPT=15740 LEN=73
Why sends the fileserver packets to this destinations?
The IP addresses are looking strange ... and the fileserver
could never get any udp packet from there because they would be dropped.
Any ideas?
Thomas.
--
--------------------------------------------------
Thomas Mueller, TU Chemnitz, URZ, D-09107 Chemnitz
--------------------------------------------------