On Mon, Jul 12, 2010 at 8:32 AM, Edward Ned Harvey <solar...@nedharvey.com>wrote:
> > From: Tim Cook [mailto:t...@cook.ms] > > > > Because VSS isn't doing anything remotely close to what WAFL is doing > > when it takes snapshots. > > It may not do what you want it to do, but it's still copy on write, as > evidenced by the fact that it takes instantaneous snapshots, and snapshots > don't get overwritten when new data is written. > > I wouldn't call that "not even remotely close." It's different, but > definitely the same ballpark. > > Everyone's SNAPSHOTS are copy on write BESIDES ZFS and WAFL's. The filesystem itself is copy-on-write for NetApp/Oracle, which is why there is no performance degradation when you take them. Per Microsoft: When a change to the original volume occurs, but before it is written to disk, the block about to be modified is read and then written to a “differences area”, which preserves a copy of the data block before it is overwritten with the change. That is exactly how pretty much everyone else takes snapshots in the industry, and exactly why nobody can keep more than a handful on disk at any one time, and sometimes not even that for data that has heavy change rates. It's not in the same ballpark, it's a completely different implementation. It's about as similar as a gas and diesel engine. They might both go in cars, they might both move the car. They aren't remotely close to each other from a design perspective. --Tim
_______________________________________________ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
