Is EMC's CAS Centera considered "permanent data"?

hashes to a specific number, but it's still not good. However,
whether it affects Centera depends on how Centera is used.

OK, this math weakness of MD5 will cause Centera to replace one piece of junk
random data without another piece of junk random data.

The probability of these MD5-breaking combinations to be of any meaning is very
small.

--
Maxim Shatskih, Windows DDK MVP
StorageCraft Corporation

http://www.storagecraft.com

"Maxim S. Shatskih" writes:
The probability of these MD5-breaking combinations to be of any
meaning is very small.

No, remember that they're deliberately constructed, and lots of file
formats have redundant or extraneous bytes in them (look how much
bigger a .doc file is than the text it contains). You can use those
extra bytes to make files collide, and use characteristics of the file
format to give the colliding files differing meanings. See:

http://www.cits.rub.de/MD5Collisions/

wrote:
: - have firmware versions the array controllers recognize (most
: high-end arrays will only accept drives of known model and firmware
: version),

This is a biggie for more reasons than you list below. The big guys,
[EHI][BDM][CMS] (interesting anagram) often have custom firmware loaded
on the drives that amongst other things include custom commands and transport
protocol variations implemented. Substituting open market drives will not
work, especially in the case of customized commands. Disk drive vendors
would have to carry these customizations forward to the newer drives in
order for them to continue to work in older kit. Now, the drive vendors
probably have to do this anyway but as we all know requirements/specifications
don't stand still over time so there are no guarantees here (as one might
be tempted to think).

Dave