Description of problem: Crash is seen in posix_getxattr (). Backtrace below: patchset: git://git.gluster.com/glusterfs.git signal received: 6 time of crash: 2012-07-04 13:42:53 configuration 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 3.3.0qa45 /lib64/libc.so.6[0x32c4632900] /lib64/libc.so.6(gsignal+0x35)[0x32c4632885] /lib64/libc.so.6(abort+0x175)[0x32c4634065] /lib64/libc.so.6[0x32c466f977] /lib64/libc.so.6[0x32c4675296] /usr/lib64/libglusterfs.so.0(data_destroy+0x55)[0x32c5a14675] /usr/lib64/libglusterfs.so.0(dict_destroy+0x36)[0x32c5a15026] /usr/lib64/glusterfs/3.3.0qa45/xlator/storage/posix.so(posix_getxattr+0x552)[0x7f86c3de8462] /usr/lib64/glusterfs/3.3.0qa45/xlator/features/access-control.so(posix_acl_getxattr+0x148)[0x7f86c3bd7c68] /usr/lib64/glusterfs/3.3.0qa45/xlator/features/locks.so(pl_getxattr+0x1a9)[0x7f86c39c37b9] /usr/lib64/glusterfs/3.3.0qa45/xlator/performance/io-threads.so(iot_getxattr_wrapper+0x122)[0x7f86c37aa5d2] /usr/lib64/libglusterfs.so.0(call_resume+0xda7)[0x32c5a2f147] /usr/lib64/glusterfs/3.3.0qa45/xlator/performance/io-threads.so(iot_worker+0x15b)[0x7f86c37b1d9b] /lib64/libpthread.so.0[0x32c4e077f1] /lib64/libc.so.6(clone+0x6d)[0x32c46e5ccd] Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. 2. 3. Actual results: Expected results: Additional info:
getxattr, removexattr and setxattr operations running parallely on a file (xattr is also same) will reproduce the crash. A patch (http://review.gluster.com/#change,3640) has been sent for this. Once it gets accepted in master, will be backported to release-3.3 branch.
This bug is not seen in current master branch (which will get branched as RHS 2.1.0 soon). To consider it for fixing, want to make sure this bug still exists in RHS servers. If not reproduced, would like to close this.
http://review.gluster.com/3673 and http://review.gluster.com/3681 fixes this bug also. Its fixed now.
CHANGE: http://review.gluster.org/4319 (cluster/afr: check for the -ve values returned from dict_serialized_length) merged in master by Anand Avati (avati)
Is not reproducible in glusterfs-3.4.0qa5. Ran a million parallel setfattr (setting and deleting xattrs) and getfattr operations. No crashes.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. http://rhn.redhat.com/errata/RHBA-2013-1262.html