Positive vm_rss: process pid 17097 comm gfs_glockd flags 400040, vma 0729d338 0804f000-080500
00 0926b4c0 rss 1 freed 0
 flags 101873, ub 100
 [<0216436f>] warn_bad_zap+0xff/0x120
 [<0216480d>] zap_page_range+0x16d/0x180
 [<02165a78>] unmap_mapping_range+0x168/0x1e0
 [<0a97327b>] gfs_inval_pte+0x7b/0xa0 [gfs]
 [<0a955160>] inode_go_drop_th+0x10/0x20 [gfs]
 [<0a950636>] rq_demote+0xb6/0xd0 [gfs]
 [<0a950092>] gfs_holder_get+0x42/0x50 [gfs]
 [<0a95077d>] run_queue+0xdd/0xf0 [gfs]
 [<0a9508db>] unlock_on_glock+0x2b/0x40 [gfs]
 [<0a9538c7>] gfs_reclaim_glock+0x157/0x1c0 [gfs]
 [<02121030>] default_wake_function+0x0/0x20
 [<0a941a28>] gfs_glockd+0x58/0x110 [gfs]
 [<02121030>] default_wake_function+0x0/0x20
 [<02121030>] default_wake_function+0x0/0x20
 [<0a9419d0>] gfs_glockd+0x0/0x110 [gfs]
 [<02104321>] kernel_thread_helper+0x5/0x14

Allows a cluster of computers to simultaneously use a block 
device that is shared between them (with FC, iSCSI, NBD, etc...). 
GFS reads and writes to the block device like a local filesystem, 
but also uses a lock module to allow the computers coordinate 
their I/O so filesystem consistency is maintained.  One of the 
nifty features of GFS is perfect consistency -- changes made to 
the filesystem on one machine show up immediately on all other 
machines in the cluster.

# cd /usr/src/redhat/SPEC
# rpmbuild --target=i686 -bb kernel-ovz.spec

# gfs_mkfs -t cluster:st01 -p lock_dlm -j 2 /dev/hdc1
...
# ccs_tool create cluster
# ccs_tool addfence manual fence_manual
# ccs_tool addnode -v 1 -f manual {hostname}
# reboot (в новое ядро 2.6.9-023stab016.2-enterprise)
...

# mkdir /vz
# mount -t gfs /dev/hdc1 /vz