On 31/03/2010 22:05, Karel Gardas wrote:
ZFS data corruption test by researchers:
http://www.cs.wisc.edu/wind/Publications/zfs-corruptio
n-fast10.pdf
Interesting paper. The only problem with it is that ZFS is filesystem and not
memory-protection vehicle. If you need memory protection use ECC or more
advanced ECC.
Well, while I know where you coming from one may argue that there is no
point in ZFS doing all the checksumming - if you need one then make sure
your underlaying protocols, buses, etc. do checksum and correct data.
But then it is probably not feasible for ZFS to double check each
block's checksum in a memory as an overhead would be too big and it
wouldn't protect an application from a data corruption entirelly anyway
as what happend if data is corrupted in memory before applications does
a write() or after it does a read() (and application buffer is corruped).
However what's good about the document is that putting memory "issue"
aside ZFS is actually shining there and passed 5/5 :)
--
Robert Milkowski
http://milek.blogspot.com
_______________________________________________
opensolaris-discuss mailing list
[email protected]
