From: "Geo." 

"Joe Barr"  wrote in message
news:pan.2003.05.10.00.42.16.44772{at}austin.rr.com...


> You skipped right over the XFS option that he suggested.

I didn't skip over it, I don't know if XFS can withstand a power cycle.

>  And you are ignoring the bigger point:  Microsoft dictated the terms of
the benchmark.

I didn't ignore that, I agreed that RAID0 could easily explain the difference.

> They selected the fastest options available for Windows and the slowest
> for Linux.  Period.

Of course, that's typical, I'm just trying to find some useful information
in these slanted results.

> And why are you afraid of Raid 5, as he suggested?

I'm not afraid of Raid5, I suggested testing it that way but I don't know
anyone running a file server with software raid5 so I don't really consider
that a good test, both with hardware raid5 perhaps would be more
meaningful. I do think software RAID1 would be a valid configuration (I run
software RAID1 on a number of servers).

Geo.

--- BBBS/NT v4.01 Flag-4