Sun today released its next generation filesystem, ZFS a few minutes ago by
pushing out the "community" (i.e. testing) build 27 of OpenSolaris in
source and binary form. You can take a look at the homepage,
documentation,
source code
tour, and a
"getting
started" screencast demo.
Thread beginning with comment 60962
To read all comments associated with this story, please click here.
To read all comments associated with this story, please click here.
Member since:>And if I read the article correctly, a form of RAID is built into the FS itself?
Yes it's called RAID-Z all details will be explained on the various SUN blogs.
The checksums you mentioned are used for the self-healing on-read feature. How I understand it only available when using some form of redundancy (mirror at least).
Now make a GPL'fied version for Linux!
Member since:2005-11-18
> > And if I read the article correctly, a form of RAID is built into the FS itself?
> Yes it's called RAID-Z all details will be explained on the various SUN blogs.
I just posted a fairly detailed description of how RAID-Z works if you're curious:
http://blogs.sun.com/roller/page/bonwick?entry=raid_z
And thank you, folks, for the kind words.
Jeff
Recent Original Stories
Recent Comments
Headlines
Random Comments
Random Stories
Random OS Link
Member since:
2005-07-06
I could do without the "and created a storage system that's actually a pleasure to use." stuff, but then again the listed features aren't bad at all!
-128 bits so that means Zettabyte capability (according to wikipedia, not the article at least)
-built in compression, which speeds up I/O (but takes up processor cycles I presume)
-Streamlined I/O all over the place
-automatically checksummed (which probably requires some extra processor cycles, but does make me a happy camper)
And if I read the article correctly, a form of RAID is built into the FS itself?
"ZFS presents a pooled storage model that completely eliminates the concept of volumes and the associated problems of partitions, provisioning, wasted bandwidth and stranded storage. Thousands of filesystems can draw from a common storage pool, each one consuming only as much space as it actually needs. The combined I/O bandwidth of all devices in the pool is available to all filesystems at all times."
This should mean it replaces the current volume system, but it's incompatible with all existing filesystems AFAICS. You can 'embed' filesystems within ZFS, but it would be using ZFS but with different policies. And it's a pretty neat system by itself after all, but interoperability needs to be taken into account.