[Tech] EIDE software RAID: montare gli HD in configurazione master/slave?

[Alessandro Zarrilli]

Vorrei configurare 4 dischi EIDE (20 GB cadauno, UDMA 100) in
modalita' RAID-5, il tutto utilizzando il software RAID di Linux. (NB:
in un precedente post chiedevo informazioni su RAID SCSI, ma fatti due
conti, non mi conviene) Stando a quanto dice il "Software-RAID HOWTO"
(http://www.linuxdoc.org/HOWTO/Software-RAID-HOWTO-3.html) sembra che
per ottenere la massima performance sia meglio montare un singolo
drive per ogni canale EIDE. Ed "a naso" mi sembra una cosa sensata:

"It is very important, that you only use one IDE disk per IDE bus. Not
only would two disks ruin the performance, but the failure of a disk
often guarantees the failure of the bus, and therefore the failure of
all disks on that bus.  In a fault-tolerant RAID setup (RAID levels
1,4,5), the failure of one disk can be handled, but the failure of two
disks (the two disks on the bus that fails due to the failure of the
one disk) will render the array unusable. Also, when the master drive
on a bus fails, the slave or the IDE controller may get awfully
confused. One bus, one drive, that's the rule."

Curiosando poi sul sito della Promise (http://www.promise.com), noti
produttori di controller EIDE, ho trovato una pagina intitolata
"Benchmarking Striped Arrays connected to the
FastTrak33/FastTrak66/FastTrak100 Controllers" (
http://www.promise.com/techsupport/benchmark/benchmarking_drives_FastT
rak33_66_100.htm) dove si consiglia, al fine di incrementare la
performance, di montare i dischi in configurazione master/slave:

"Create Two Drive Stripes with    the drives configured as Master &
Slave on the IDE Channel 1 or 2, but    not both. This can help reduce
PCI BUS Latency therefore increase the    Performance of the FastTrak
Controllers"

Adesso io mi domando, a parte il fatto che "the failure of a disk
often guarantees the failure of the bus, and therefore the failure of
all disks on that bus", per quanto riguarda il throughput, chi ha
ragione?