Gluster users - Apr 2009 - loop mount makes the system hang and reboot

Hi,
 
I tried to use Gluster 1.3.12. The installation works fine, the remote files can
be create and accessed normally.
 
But when I try to loop mount a file(8G image), it will hang the system and makes
it reboot. (loop mount a local file is ok)
 
Gluster version is 1.3.12
Fuse, tried 2.7.4 and Gluster patched 2.7.4
Kernel is 2.6.18-8.1.15.1.16.el5xen  (using Oracle Virtual Server, which is
essentially Xen 3.1.4)
 
I specified -d disable when launching glusterfs on client
Mount command is:  mount -o loop /path/to/file /mnt
 
Last few lines of gluster log on client.
2009-04-22 05:08:28 D [fuse-bridge.c:529:fuse_lookup] glusterfs-fuse: 46: LOOKUP
/dkt_disk1(14516228)
2009-04-22 05:08:28 D [fuse-bridge.c:384:fuse_entry_cbk] glusterfs-fuse: 46:
(34) /dkt_disk1 => 14516228
2009-04-22 05:08:28 D [inode.c:367:__active_inode] fuse/inode: activating
inode(14516228), lru=0/1024
2009-04-22 05:08:28 D [fuse-bridge.c:1553:fuse_open] glusterfs-fuse: 47: OPEN
/dkt_disk1
2009-04-22 05:08:28 D [fuse-bridge.c:687:fuse_fd_cbk] glusterfs-fuse: 47: (12)
/dkt_disk1 => 0x99ec2e8

 
last few lines of gluster log on server (2 servers, in unify)
 
2009-04-22 21:40:17 D [inode.c:367:__active_inode] brick-ns/inode: activating
inode(14516227), lru=1/1024
2009-04-22 21:40:17 D [inode.c:367:__active_inode] brick/inode: activating
inode(14516226), lru=0/1024
2009-04-22 21:40:17 D [inode.c:367:__active_inode] brick-ns/inode: activating
inode(14516228), lru=1/1024
2009-04-22 21:40:23 D [inode.c:367:__active_inode] brick-ns/inode: activating
inode(14516228), lru=1/1024
2009-04-22 21:40:23 D [inode.c:367:__active_inode] brick-ns/inode: activating
inode(14516228), lru=1/1024
 
Any suggestion how to solve this issue?
 
Thanks,
Frank

 
 
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://supercolony.gluster.org/pipermail/gluster-users/attachments/20090422/66c2aa3f/attachment.html>