Bug 1043862

Description of problem:
=======================

While trying to create snapshot(same or different snap name ) for the same volume from all the nodes in cluster, glusterd crashed on all the nodes in cluster with the following logs:

[2013-12-17 01:27:59.578772] E [glusterd-locks.c:233:glusterd_volume_lock] 0-: Unable to acquire lock. Lock for vol-snap0 held by 8795e6ed-a3b7-4c04-b397-d938a05dd1d4
[2013-12-17 01:27:59.578825] E [glusterd-locks.c:174:glusterd_multiple_volumes_lock] 0-: Failed to acquire lock for vol-snap0. Unlocking other volumes locked by this transaction
[2013-12-17 01:27:59.578842] E [glusterd-mgmt-handler.c:85:glusterd_syctasked_volume_lock] 0-: Failed to acquire volume locks on localhost
[2013-12-17 01:27:59.579111] E [rpcsvc.c:495:rpcsvc_check_and_reply_error] 0-rpcsvc: rpc actor failed to complete successfully
2013-12-17 01:27:59.579670] E [glusterd-mgmt.c:108:gd_mgmt_v3_collate_errors] 0-: Locking volume failed on 10.70.43.20. Please check log file for details.
pending frames:
frame : type(0) op(0)
frame : type(0) op(0)
frame : type(0) op(0)
frame : type(0) op(0)

patchset: git://git.gluster.com/glusterfs.git
signal received: 6
time of crash: 2013-12-17 01:27:59configuration details:
argp 1
backtrace 1
dlfcn 1
fdatasync 1
libpthread 1
llistxattr 1
setfsid 1
spinlock 1
epoll.h 1
xattr.h 1
st_atim.tv_nsec 1
package-string: glusterfs 3.4.0.snap.dec03.2013git
/lib64/libc.so.6[0x312ea329a0]
/lib64/libc.so.6(gsignal+0x35)[0x312ea32925]
/lib64/libc.so.6(abort+0x175)[0x312ea34105]
/lib64/libc.so.6[0x312ea70837]
/lib64/libc.so.6[0x312ea76166]
/lib64/libc.so.6[0x312ea79f8a]
/lib64/libc.so.6(__libc_calloc+0xc6)[0x312ea7a626]
/usr/lib64/libglusterfs.so.0(__gf_calloc+0x53)[0x7f0fd92e5b33]
/usr/lib64/libglusterfs.so.0(iobref_new+0x15)[0x7f0fd92e6cd5]
/usr/lib64/glusterfs/3.4.0.snap.dec03.2013git/rpc-transport/socket.so(+0x944c)[0x7f0fd51dc44c]
/usr/lib64/glusterfs/3.4.0.snap.dec03.2013git/rpc-transport/socket.so(+0xa7bd)[0x7f0fd51dd7bd]
/usr/lib64/libglusterfs.so.0(+0x66097)[0x7f0fd930c097]
/usr/sbin/glusterd(main+0x53a)[0x40680a]
/lib64/libc.so.6(__libc_start_main+0xfd)[0x312ea1ed1d]
/usr/sbin/glusterd[0x404629]
---------

========================================================================

Following set of commands were in execution from all the nodes in cluster simultaneously

gluster snapshot create vol-snap0 -n snap3
gluster snapshot create vol-snap0 -n snap4
gluster snapshot create vol-snap0 -n snap5
gluster snapshot create vol-snap0 -n snap6
gluster snapshot create vol-snap0 -n snap7

expected result was:
===================

1. If a snap already exist, fail the command with message "snap already exist"
2. If a snap name doesnt exist than acquire a lock from one machine on a volume and success the snap creation.

actual result:
===============

glusterd crashed on all the nodes in cluster



Version-Release number of selected component (if applicable):
=============================================================

glusterfs-3.4.0.snap.dec03.2013git-1.el6.x86_64