zfs

#CREATING
- you can create a zpool manager through img files just like btrfs but its not recommended as it looses all of zfs' control of underlying disks and resiliance depends on the underlying fs. But thats ok since we only care about clone speeds and ability to scale
creating/destroying new filesystems.
- the mount table is populated automatically, will be interesting to how it scales on 100s of filesystems
- These are pretty avg times
- if this is how fast new creates take, damn those cow clones gunna blaze it
root@ownedbykrishna:/home/kmannem# time zfs create test/diek
real    0m0.033s
user    0m0.002s
sys     0m0.013s
- you can create sub zfs filesystems under a root, you cant delete the root unless all the children are gone
zfs create bigone/biggie
zfs create bigone/biggie/smallie

# DESTROYING
- mounts are automatically removed from the mtab
- i think the destroys depend on how big the actual fs is. This is what an empty one takes. 
root@ownedbykrishna:/home/kmannem# time zfs destroy test/hello
real    0m0.035s
user    0m0.005s
sys     0m0.002s
- this is how long a fs with 30MB takes
root@ownedbykrishna:/# time zfs destroy test/home
real    0m0.074s
user    0m0.004s
sys     0m0.005s
- even if its big, ~1GB its coolio
root@ownedbykrishna:/# time zfs destroy bigone/biggie
real    0m0.060s
user    0m0.000s
sys     0m0.012s


#CLONING