RE: IDE RAID support -- FRAID v. intelligent hardware RAID

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



Nick Bryant <nick@xxxxxxxxxxxxxxxxxxxxxx> wrote:
> Indeed. I freaked out when I couldn't get my sata promise
> raid controlling working in Acer Altos series servers...
> Till I read an article that said in 2.4 its actually just
> as quick (if not quicker, depending on server hardware)
> to let the kernel deal with the raid stuff.

The Logical Disk/Volume Management (LDM/LVM) of the OS is
always going to order, schedule and commit far more efficiently
than a FRAID driver (FRAID cards have *0* intelligence, only
a "trick" 16-bit BIOS that is _useless_ once the OS boots).

Software RAID-0 may be faster than hardware (especially
if spread over cards on separate PCI channels).

Software RAID-1 and 10 require the I/O commit 2x the data
over hardware RAID, which just takes the 1:1 stream.

Software RAID-3/4/5 push _all_ data through the CPU-memory-
I/O interconnect, and is detrimental on the limited CPU-
memory-I/O interconnect (the XORs aren't the problem),
whereas hardware RAID-3/4/5 just takes the 1:1 stream.


-- 
Bryan J. Smith                 mailto:b.j.smith@xxxxxxxx
Sent from Yahoo Mail (please excuse any missing headers)

[Index of Archives]     [CentOS]     [CentOS Announce]     [CentOS Development]     [CentOS ARM Devel]     [CentOS Docs]     [CentOS Virtualization]     [Carrier Grade Linux]     [Linux Media]     [Asterisk]     [DCCP]     [Netdev]     [Xorg]     [Linux USB]
  Powered by Linux