On Tue, Mar 10, 2009 at 6:04 PM, Alan Altmark <[email protected]> wrote:

> For large-scale sharing, xip2fs and DCSSes are the only way (IMO) to share
> files in the way we all originally envisioned minidisk sharing.  The DCSS
> owner makes all the updates and then saves a new copy of the DCSS.  New
> users get the new one - previous users still have the old DCSS.  When no
> one is using the old one any more, it magically disappears.  And the
> DCSSes always have consistent content, even if someone loads the DCSS
> while you're in the middle of building a new one.

And you will find yourself in a pile of hardly shared DCSS files that
gobble up storage...

The atomic operation is also nicely implemented by a pool of mini
disks. DDR the latest to the next unused disk, do your updates, umount
and get rid of your R/W link, rotate the directory entries so new
users pick up the next generation. The next server that tries to link
the common R/O disk gets the new version, and eventually all links to
the old disk will appear as by magic ;-)

My crystal ball has a stack of sparse file systems: a large R/O disk,
a DCSS, and one or more private R/W disks for the servers.

Rob

----------------------------------------------------------------------
For LINUX-390 subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390

Reply via email to