[Mauiusers] Maui crashes: "double free or corruption (out)"

Klauninger Bert Bert.Klauninger at arcs.ac.at
Tue Sep 9 00:38:41 MDT 2008


Dear list,

 

I was successful in compiling MAUI 3.2.6p19 on an AMD Opteron 252 -
operating system is Ubuntu Server 8.04 x86_64, PBS is Torque 2.3.3. From
time to time (usually once a day on our HPC Cluster which is under heavy
usage currently) the maui process crashes with the following output:

 

*** glibc detected *** /usr/local/maui/sbin/maui: double free or
corruption (out): 0x00000000022f4060 ***

======= Backtrace: =========

/lib/libc.so.6[0x7f9b8fe3108a]

/lib/libc.so.6(cfree+0x8c)[0x7f9b8fe34c1c]

/lib/libc.so.6(_IO_free_backup_area+0x18)[0x7f9b8fe2d5c8]

/lib/libc.so.6(_IO_file_close_it+0x32)[0x7f9b8fe2bd02]

/lib/libc.so.6(fclose+0x19a)[0x7f9b8fe1fdfa]

/usr/local/maui/sbin/maui[0x45a631]

/usr/local/maui/sbin/maui[0x40387e]

/usr/local/maui/sbin/maui[0x405ce4]

/lib/libc.so.6(__libc_start_main+0xf4)[0x7f9b8fddb1c4]

/usr/local/maui/sbin/maui[0x403099]

======= Memory map: ========

00400000-00500000 r-xp 00000000 08:01 1652763
/usr/local/maui/sbin/maui

00700000-00705000 rw-p 00100000 08:01 1652763
/usr/local/maui/sbin/maui

00705000-02473000 rw-p 00705000 00:00 0
[heap]

7f9b88000000-7f9b88021000 rw-p 7f9b88000000 00:00 0

7f9b88021000-7f9b8c000000 ---p 7f9b88021000 00:00 0

7f9b8f9a3000-7f9b8f9b0000 r-xp 00000000 08:01 1488113
/lib/libgcc_s.so.1

7f9b8f9b0000-7f9b8fbb0000 ---p 0000d000 08:01 1488113
/lib/libgcc_s.so.1

7f9b8fbb0000-7f9b8fbb1000 rw-p 0000d000 08:01 1488113
/lib/libgcc_s.so.1

7f9b8fbb1000-7f9b8fbbb000 r-xp 00000000 08:01 1489136
/lib/libnss_files-2.7.so

7f9b8fbbb000-7f9b8fdbb000 ---p 0000a000 08:01 1489136
/lib/libnss_files-2.7.so

7f9b8fdbb000-7f9b8fdbd000 rw-p 0000a000 08:01 1489136
/lib/libnss_files-2.7.so

7f9b8fdbd000-7f9b8ff15000 r-xp 00000000 08:01 1489127
/lib/libc-2.7.so

7f9b8ff15000-7f9b90115000 ---p 00158000 08:01 1489127
/lib/libc-2.7.so

7f9b90115000-7f9b90118000 r--p 00158000 08:01 1489127
/lib/libc-2.7.so

7f9b90118000-7f9b9011a000 rw-p 0015b000 08:01 1489127
/lib/libc-2.7.so

7f9b9011a000-7f9b9011f000 rw-p 7f9b9011a000 00:00 0

7f9b9011f000-7f9b90152000 r-xp 00000000 08:01 1328603
/usr/local/lib/libtorque.so.2.0.0

7f9b90152000-7f9b90352000 ---p 00033000 08:01 1328603
/usr/local/lib/libtorque.so.2.0.0

7f9b90352000-7f9b90355000 rw-p 00033000 08:01 1328603
/usr/local/lib/libtorque.so.2.0.0

7f9b90355000-7f9b90380000 rw-p 7f9b90355000 00:00 0

7f9b90380000-7f9b90400000 r-xp 00000000 08:01 1489131
/lib/libm-2.7.so

7f9b90400000-7f9b905ff000 ---p 00080000 08:01 1489131
/lib/libm-2.7.so

7f9b905ff000-7f9b90601000 rw-p 0007f000 08:01 1489131
/lib/libm-2.7.so

7f9b90601000-7f9b9061e000 r-xp 00000000 08:01 1489124
/lib/ld-2.7.so

7f9b9062f000-7f9b9076e000 rw-p 7f9b9062f000 00:00 0

7f9b9076e000-7f9b907a3000 r--s 00000000 08:01 1275711
/var/cache/nscd/hosts

7f9b907a3000-7f9b907d8000 r--s 00000000 08:01 1276450
/var/cache/nscd/group

7f9b907d8000-7f9b9080d000 r--s 00000000 08:01 1276444
/var/cache/nscd/passwd

7f9b9080d000-7f9b9080f000 rw-p 7f9b9080d000 00:00 0

7f9b9081a000-7f9b9081e000 rw-p 7f9b9081a000 00:00 0

7f9b9081e000-7f9b90820000 rw-p 0001d000 08:01 1489124
/lib/ld-2.7.so

7fff98590000-7fff9881f000 rw-p 7fffffd70000 00:00 0
[stack]

7fff989b9000-7fff989bb000 r-xp 7fff989b9000 00:00 0
[vdso]

ffffffffff600000-ffffffffff601000 r-xp 00000000 00:00 0
[vsyscall]

 

My maui.cfg file looks like this:

 

SERVERHOST             mash

ADMIN1                 root

RMCFG[MASH]            TYPE=PBS

#RMPOLLINTERVAL        00:00:30

 

#SERVERPORT            42559

#SERVERMODE            NORMAL

 

LOGFILE                maui.log

LOGFILEMAXSIZE         100000

LOGLEVEL               7

 

#QUEUETIMEWEIGHT        1

 

# default anyway

#BACKFILLPOLICY         FIRSTFIT

#RESERVATIONPOLICY      CURRENTHIGHEST

 

#NODEACCESSPOLICY       SHARED

#NODEALLOCATIONPOLICY   LASTAVAILABLE

 

#PREEMPTPOLICY          SUSPEND

 

CLASSWEIGHT            1

QOSWEIGHT              1

 

CLASSCFG[gromacs]       QDEF=med

CLASSCFG[blast]         QDEF=hi

CLASSCFG[lazy]          QDEF=low

CLASSCFG[amber]         QDEF=med

 

QOSCFG[hi]              PRIORITY=100000 QFLAGS=PREEMPTOR FLAGS=PREEMPTOR

QOSCFG[med]             PRIORITY=10000 QFLAGS=PREEMPTOR FLAGS=PREEMPTOR

QOSCFG[low]             PRIORITY=100 QFLAGS=PREEMPTEE FLAGS=PREEMPTEE

 

 

I could not find a bug tracker on supercluster.org, so I'm posting this
here. Has anyone of you had similar effects?

 

Best regards

Bert

 

DI Bert Klauninger
Austrian Research Centers GmbH - ARC
Biogenetics/Natural Resources
2444 Seibersdorf
Austria
phone:  +43 (0)50550-3630
fax:      +43 (0)50550-3666
www.arcs.ac.at <http://www.arcs.ac.at/> 
www.picme.at <http://www.picme.at/> 

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.supercluster.org/pipermail/mauiusers/attachments/20080909/84f58211/attachment-0001.html


More information about the mauiusers mailing list