[ILUG] Highpoint Sata RAID

[Gareth Eason]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


	Are these the special 'RAID' cards that store some config information
on a flash chip on the card - so if your RAID card dies, you've just
lost your entire dataset... because even if you purchase exactly the
same card there's no way to configure it in exactly the same way?

	Or are those different 'RAID' cards I'm thinking of?

	Best regards,
	-->Gar



Paul Jakma wrote:
| On Wed, 23 Mar 2005, Niall O Broin wrote:
|
|> Hardware assisted RAID maybe? But still, host drivers needed. Let me
|> modify my definition of hardware RAID then - hardware RAID is where
|> the operating system has NO involvment in doing the actual RAID work
|> (it will of course have to have drivers for the card) and the OS sees
|> as disks only the RAID volumes provided by the controller, no matter
|> what the physical disk layout is.
|
|
| It's as good as hardware RAID. The performance sensitive stuff is the
| bus copy (pure soft RAID needs more bus bandwidth) and any ECC (XOR for
| RAID-5 usually, afaik). If that performance sensitive stuff is on the
| card, who cares whether management is done on a card CPU or a host CPU?
| Same difference performance wise (and i'd rather have it in my OS where
| it'll get broader testing).
|
|> So you can use the not very well tested Promise Linux driver (if there
|> is one for that card) or md without hardware assistance. I know which
|> I'd use.
|
|
| Indeed, but MD will gain support for offload in time.
|
| Wasn't arguing for the use of some weird oddball RAID driver. Just
| pointing out the promise SX4 /hardware/ is very interesting, and as good
| as hardware raid (even if linux MD doesnt yet make use of it).
|
| regards,
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (MingW32)

iD8DBQFCQYndK36C50PvIR8RApNxAJ93iAgtBuWeBGNvqS3VDvG3tFTTvQCfccYh
q/xtd07IxyyhO6JgEsSb5OQ=
=u51O
-----END PGP SIGNATURE-----