ZFS: what’s not to like?

I can not wait for the some form of ZFS to hit the Windows arena. For the uninitiated ZFS is Sun Microsystem’s new filesystem, that is simply better (on paper) than anything else. Period.

What is this guy on about?

Well, with a ZFS filesystem, you can have:

• Very simple administration.
• Simply massive volume sizes.
• Self healing: end to end checksums recover bad blocks – automatically, seamlessly in the background!!
• Dynamic capacity increases: on a live production system, you can simply issue a single command to grow the disk space.
• RAIDZ – RAID5 replacement without the write-hole issues.

There are other virtues, but for me, the real killer is that it’s a self-healing, error correcting filesystem. For years, as a photographer, I have had to create very elaborate systems to overcome ‘bit-rot’ (slow corruption of data over time) wreaking havoc with my images.

ZFS appears to have solved this issue in a single stroke.

Moreover, I could add extra storage to the ZFS ‘pool’ easily with a  simple command. Try doing that on a RAID5.

However, ZFS is only really fully implemented on Solaris – an operating system I have little experience with – worse my entire workflow is based upon Windows applications…so I guess I will have to wait for Microsoft to come up with a ZFS alternative.

Or switch to Solaris?