[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: Fedora 9 Boot Problem

Seann Clark wrote:
> under linux, with the HP370 driver loaded into the kernel, it will work  
> as a RAID that isn't managed or taxing the CPU. The HP controller will  
> take care of all mirroring and striping aspects leaving the CPU free to  
> do whatever else.

Sorry, but I’m pretty sure this is incorrect.

It handles the RAID transparently to the system, that’s true – but it
uses the system CPU. The RAID is done *in* the HPT 370 driver.

Fortunately, this doesn’t take much of the CPU’s time for the RAID modes
the HPT 370 supports – they’re not processor-intensive.

Actually finding documentary evidence for this seems to be rather
difficult: Highpoint obviously want to convince people that it’s Real
RAID (honest!), and the Linux kernel world seems to just assume that
it’s all done in BIOS and drivers. The best quote I can find after a bit
of Googling and grepping is:
where Alan Cox (in a thread called “binary kernel drivers re. hpt370 and
redhat” (sic) says
    Pretty much all of the older PATA controllers don’t actually do
    hardware raid but bios/driver raid - ie its the equivalent (or
    roughly so) of the md layer but locks you into the vendor. The
    notable exception here is the 3ware card (there are a
    couple of others too - Promise Supertrak100, SX6000)

    We know some of these formats (eg see the hptraid driver in 2.4.2x) 

Incidentally, I find the name “fake raid” fun, but unfortunate: it is
still real RAID. Just not real hardware RAID.

Hope this helps,


E-mail:     james@ | Which do you consider was the stronger swimmer,
aprilcottage.co.uk |     (a) The Spanish Armadillo,
                   |     (b) The Great Seal?
                   |     -- ‘1066 and All That’

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]