Hi:
I am installing lustre 1.5.95 on a small sun x4200 cluster running
FC5 kernel version 2.6.15. I had some difficulty getting kernel
patched and rpms compiled, but by reading discussion i was able to
get over those problems. The problem now is i am testing the simplest
scenario and it failes. I put mgs and mdt on node 1, called gg202.
OST on the second node, called gg203. Then client on the third node,
called gg204. All the mkfs and mount on these three nodes seems fine,
and /proc/fs/lustre/devices seems ok as well. But after i mount the
client at gg204 when try to copy sth to the mounted directory on
gg204, the gg202 node, i.e., the mgs and mdt node, stops responding.
I have to manually power it down. I can not find any useful
information at gg201''s system log, and gg202 and gg203''s
system log
simply says connection lost. So now i am really confused. Thanks for
the patience.
Dayong Huang
Center of Computation and Technology
Louisiana State University
On the MDT MGS node, gg202:
[root@gg202]# mkfs.lustre --fsname=mainfs --mdt --mgs --reformat /dev/
VolGroup00/MDT
Permanent disk data:
Target: mainfs-MDTffff
Index: unassigned
UUID:
Lustre FS: mainfs
Mount type: ldiskfs
Flags: 0x75
(MDT MGS needs_index first_time update )
Persistent mount opts: errors=remount-ro,iopen_nopriv,user_xattr
Parameters:
device size = 224MB
formatting backing filesystem ldiskfs on /dev/VolGroup00/MDT
target name mainfs-MDTffff
4k blocks 0
options -i 4096 -I 512 -q -O dir_index -F
mkfs_cmd = mkfs.ext2 -j -b 4096 -L mainfs-MDTffff -i 4096 -I 512 -q -
O dir_index -F /dev/VolGroup00/MDT
Writing CONFIGS/mountdata
[root@gg202]# mount -t lustre /dev/VolGroup00/MDT /mnt/mainfs/mdt/
[root@gg202]# cat /proc/fs/lustre/devices
0 UP mgs MGS MGS 5
1 UP mgc MGC192.168.1.102@tcp fe3d6e35-83cb-93bf-ea9d-ac3bfa796a6f 5
2 UP mdt MDS MDS_uuid 3
3 UP lov mainfs-mdtlov mainfs-mdtlov_UUID 4
4 UP mds mainfs-MDT0000 mainfs-MDT0000_UUID 3
On the OST node, gg203:
[root@gg203]# mkfs.lustre --fsname=mainfs --ost --mgsnode=gg202@tcp0
--reformat /dev/VolGroup00/OST2
Permanent disk data:
Target: mainfs-OSTffff
Index: unassigned
UUID:
Lustre FS: mainfs
Mount type: ldiskfs
Flags: 0x72
(OST needs_index first_time update )
Persistent mount opts: errors=remount-ro,extents,mballoc
Parameters: mgsnode=192.168.1.102@tcp
device size = 20480MB
formatting backing filesystem ldiskfs on /dev/VolGroup00/OST2
target name mainfs-OSTffff
4k blocks 0
options -J size=400 -i 16384 -I 256 -q -O dir_index -F
mkfs_cmd = mkfs.ext2 -j -b 4096 -L mainfs-OSTffff -J size=400 -i
16384 -I 256 -q -O dir_index -F /dev/VolGroup00/OST2
Writing CONFIGS/mountdata
[root@gg203]# mount -t lustre /dev/VolGroup00/OST2 /mnt/mainfs/ost2/
[root@gg203]# cat /proc/fs/lustre/devices
0 UP mgc MGC192.168.1.102@tcp 3a498b4e-884e-ed25-ee0d-2eadd19789bc 5
1 UP ost OSS OSS_uuid 3
2 UP obdfilter mainfs-OST0000 mainfs-OST0000_UUID 7
On the client node, gg204:
[root@gg204]# mount -t lustre gg202@tcp0:/mainfs /mnt/testfs/
[root@gg204]# df
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/mapper/VolGroup00-rootlv
39676508 9884128 27744380 27% /
/dev/sda1 194442 25009 159394 14% /boot
tmpfs 4021556 0 4021556 0% /dev/shm
/dev/mapper/VolGroup00-homelv
20642428 2754120 16839732 15% /home
192.168.1.102@tcp:/mainfs
20642428 455200 19138588 3% /mnt/testfs
[root@gg204]# cp /tmp/testfile /mnt/testfs/.
This call never returns and gg202 crashes.