Dear all,
Since our latest CE upgrade (RPM versions are below (*)), we find weird
addresses/hostnames (non-existing) in torque logs under the server_logs
dir (one file per day) . These hosts appear as submitters of qstat
queries, when in reality those queries come from the same CE
(localhost). Queries from other hosts (WNs e.g.) appear in general with
the correct hostname. I have done an ltrace of the pbs_server and seems
that it gets the right name of the CE (lcg02.ciemat.es) but logs
something different (in the example below (**) is 32.7.174.0). This
address/host is not always the same. There are four or five different
ones, all of them quite strange.
I would think there is some bug in the pbs code that corrupts the
requesting host, but before submitting a bug, I thought it would be good
to know if other sites see the same thing. Is that the case? Does anyone
have further ideas?
Thank you.
Cheers,
Antonio
* RPMs in our CE
-------------------------------------------------
lcg-CE-3.1.16-0
torque-server-2.3.0-snap.200801151629.2cri.slc4
torque-client-2.3.0-snap.200801151629.2cri.slc4
torque-2.3.0-snap.200801151629.2cri.slc4
torque-mom-2.3.0-snap.200801151629.2cri.slc4
glite-yaim-torque-utils-4.0.2-2
glite-yaim-torque-client-4.0.1-1
glite-yaim-torque-server-4.0.1-5
-------------------------------------------------
** ltrace
-------------------------------------------------
5767 recvmsg(12, 0xbfffdf50, 0, 0xbfffdf44, 4) = 1
5767 malloc(12) = 0x85d1d48
5767 getpwuid(42000, 0xbfffdf50, 0, 0xbfffdf44, 4) = 0x34909c
5767 strcpy(0x8093f08, "cmsprd") = 0x8093f08
5767 strcpy(0x8093f18, "lcg02.ciemat.es") = 0x8093f18
5767 free(0x85d1d48) = <void>
5767 DIS_tcp_setup(12, 13991, 0xbfffdfa8, 0x805b892, 0x8093ad0) = 0xb7f31014
5767 decode_DIS_ReqHdr(12, 0x86ee128, 0xbfffdf94, 0xbfffdf98, 0x8093ad0) = 0
5767 decode_DIS_Status(12, 0x86ee128, 0xbfffdf94, 0xbfffdf98, 0x8093ad0) = 0
5767 decode_DIS_ReqExtend(12, 0x86ee128, 0xbfffdf94, 0xbfffdf98,
0x8093ad0) = 0
5767 get_connecthost(12, 0x86ee161, 64, 0x808e3b0, 0x80c0f69) = 0
5767 reqtype_to_txt(19, 0x86ee161, 63, 0x808e3b0, 0x80c0f69) = 0xacb0bd
5767 sprintf("Type StatusJob request received from [log in to unmask],
sock=12", "Type %s request received from %s@%s, sock=%d", "StatusJob",
"cmsprd", "32.7.174.0", 12) = 63
5767 log_event(256, 4, 0x807c8fa, 0x808e3b0, 0x86ee161) = 0
-------------------------------------------------
|