[Pacemaker] crm_mon -W crash

Kazunori INOUE inouekazu at intellilink.co.jp
Wed Nov 28 08:54:56 UTC 2012


Hi,

I try to handle SNMP trap with crm_mon.
However, crm_mon crashes with SIGSEGV at the time of fencing.

[environment]
   - Red Hat Enterprise Linux Server release 6.3 (Santiago)
   - ClusterLabs/pacemaker  9c13d14640(Nov 27, 2012)
   - corosync               92e0f9c7bb(Nov 07, 2012)

   [root at dev1 ~]$ pacemakerd -F
   Pacemaker 1.1.8 (Build: 9c13d14)
    Supporting:  generated-manpages agent-manpages ascii-docs publican-docs ncurses libqb-logging libqb-ipc lha-fencing  corosync-native snmp

   [root at dev1 ~]$ rpm -qi net-snmp-libs
   Name        : net-snmp-libs              Relocations: (not relocatable)
   Version     : 5.5                             Vendor: Red Hat, Inc.
   Release     : 41.el6                      Build Date: Fri May 18 19:20:24 2012
   Install Date: Mon Jul  2 14:15:53 2012       Build Host: x86-003.build.bos.redhat.com
   -snip-


[test case]
1. set only STONITH resources and perform crm_mon with -W option.

    [root at dev1 ~]$ crm_mon -S 192.168.133.148 -W
    Last updated: Wed Nov 28 11:44:46 2012
    Last change: Wed Nov 28 11:44:35 2012 via cibadmin on dev1
    Stack: corosync
    Current DC: dev1 (2506467520) - partition with quorum
    Version: 1.1.8-9c13d14
    2 Nodes configured, unknown expected votes
    2 Resources configured.


    Online: [ dev1 dev2 ]

    prmStonith1     (stonith:external/libvirt):     Started dev2
    prmStonith2     (stonith:external/libvirt):     Started dev1

2. fence the dev2.

    [root at dev1 ~]$ crm node fence dev2
    Do you really want to shoot dev2? y

3. then, crm_mon crashed.

    [root at dev1 ~]$ crm_mon -S 192.168.133.148 -W
    Last updated: Wed Nov 28 11:45:32 2012
    Last change: Wed Nov 28 11:44:35 2012 via cibadmin on dev1
    Stack: corosync
    Current DC: dev1 (2506467520) - partition WITHOUT quorum
    Version: 1.1.8-9c13d14
    2 Nodes configured, unknown expected votes
    2 Resources configured.


    Node dev2 (2472913088): UNCLEAN (offline)
    Online: [ dev1 ]

    prmStonith1     (stonith:external/libvirt):     Started dev2
    prmStonith2     (stonith:external/libvirt):     Started dev1
    Segmentation fault (core dumped)
    [root at dev1 ~]$


GDB shows this:
   [root at dev1 ~]$ gdb `which crm_mon` core.28326
   GNU gdb (GDB) Red Hat Enterprise Linux (7.2-56.el6)
   -snip-
   Core was generated by `crm_mon -S 192.168.133.148 -W'.
   Program terminated with signal 11, Segmentation fault.
   #0  0x0000003f808805a1 in __strlen_sse2 () from /lib64/libc.so.6
   -snip-
   (gdb) bt
   #0  0x0000003f808805a1 in __strlen_sse2 () from /lib64/libc.so.6
   #1  0x0000003f81c39481 in snmp_add_var () from /usr/lib64/libnetsnmp.so.20
   #2  0x00000000004099ba in send_snmp_trap (node=0x24408a0 "dev2", rsc=0x0, task=0x245c850 "st_notify_fence", target_rc=0, rc=0, status=0, desc=0x2462b80 "Operation st_notify_fence requested by dev1 for peer dev2: OK (ref=c520e07b-907f-48b9-a216-4786289b61da)") at crm_mon.c:1716
   #3  0x000000000040af6b in mon_st_callback (st=0x2409520, e=0x243aa30) at crm_mon.c:2241
   #4  0x00007fc23639598d in stonith_send_notification (data=0x2408390, user_data=0x7fff159e1410) at st_client.c:1960
   #5  0x000000364263688c in g_list_foreach () from /lib64/libglib-2.0.so.0
   #6  0x00007fc236396638 in stonith_dispatch_internal (buffer=0x2429a08 "<notify t=\"st_notify\" subt=\"st_notify_fence\" st_op=\"st_notify_fence\" st_rc=\"0\"><st_calldata><st_notify_fence state=\"2\" st_rc=\"0\" st_target=\"dev2\" st_device_action=\"reboot\" st_delegate=\"dev1\" st_remote"..., length=387, userdata=0x2409520) at st_client.c:2128
   #7  0x00007fc235d03391 in mainloop_gio_callback (gio=0x2433fe0, condition=G_IO_IN, data=0x240a2e0) at mainloop.c:565
   #8  0x0000003642638f0e in g_main_context_dispatch () from /lib64/libglib-2.0.so.0
   #9  0x000000364263c938 in ?? () from /lib64/libglib-2.0.so.0
   #10 0x000000364263cd55 in g_main_loop_run () from /lib64/libglib-2.0.so.0
   #11 0x0000000000404e23 in main (argc=4, argv=0x7fff159e1778) at crm_mon.c:590
   (gdb)

Is this a known issue?

Best Regards,
Kazunori INOUE
-------------- next part --------------
property no-quorum-policy="ignore" \
  stonith-enabled="true" \
  startup-fencing="false" \
  stonith-timeout="60s"

primitive prmStonith1 stonith:external/libvirt \
  params \
    hostlist="dev1" \
    hypervisor_uri="qemu+ssh://bl460g1n5/system"
primitive prmStonith2 stonith:external/libvirt \
  params \
    hostlist="dev2" \
    hypervisor_uri="qemu+ssh://bl460g1n6/system"

location rscloc-prmStonith1 prmStonith1 \
  rule -INFINITY: #uname eq dev1
location rscloc-prmStonith2 prmStonith2 \
  rule -INFINITY: #uname eq dev2



More information about the Pacemaker mailing list