http://www.osrg.net/sheepdog/

Taobao.com

/etc/init.d/sheepdog start        <-- TCP / 7000

mount -o remount,user_xattr /var/lib/sheepdog
sheep /var/lib/sheepdog         <-- store objects

collie cluster format --copies=3     <--- Backend: Simple Store (default)
                                                         Do never kill more than X-1 sheep daemons

collie node list

qemu-img create sheepdog:IMG-NAME 256G
qemu-img convert -t writethrough ~/amd64.raw sheepdog:IMG-NAME
qemu-img create -b sheepdog:Alice:1 sheepdog:Charlie

collie vdi list   <--- list snapshot also

qemu-system-x86_64 sheepdog:IMG-NAME
qemu-system-x86_64 sheepdog:192.168.10.2:7000:Alice
qemu-system-x86_64 -drive file=sheepdog:Alice,cache=writeback

qemu-img snapshot -c name sheepdog:Alice
qemu-system-x86_64 sheepdog:Alice:1

collie cluster shutdown

sheepfs

sheepfs /your/mountpoint
cat sheepfs_dir/cluster/info
echo test > sheepfs_dir/vdi/mount
mkfs.ext4 sheepfs_dir/volume/test
mount -o loop sheepfs_dir/volume/test /somewhere
echo ip:port > sheepfs_dir/config/sheep_info

start multiple 'sheep' processes to simulate the cluster.

$ for i in 0 1 2; do sheep -c local -d /path/to/store/$i -z $i -p 700$i;sleep 1;done

Corosync

only work reliably with less than 15 nodes

sheep -w 100 /path/to/sheep
max cache size to 100M and use default writethrough mode