30 Apr
2014
30 Apr
'14
4:10 a.m.
On Tue, Apr 29, 2014 at 08:53:48PM +1000, Les Kitchen (LUV) wrote:
But building on this idea, you could image a scheme where you keep block-checksums (for some reasonable size of block) for a device, and when the time came, just transfer changed blocks.
this is roughly how 'zfs send' (and 'btrfs send') works. in zfs & btrfs, a snapshot is just a list of blocks, so when you do an incremental zfs send, the list of blocks for new-snapshot@localhost are compared to old-snapshote@remotehost and only the blocks that differ need be sent. it's far more efficient even than rsync (because there's no need to compare the data in the blocks between src and dst). craig -- craig sanders <cas@taz.net.au> BOFH excuse #438: sticky bit has come loose