SATA Drives, non RAID

[Guest]

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

"Milleron" <millerdot90@SPAMlessosu.edu> wrote in message
news:13rlg0tupj2d8ii3gvmafvmbft9s4ancdl@4ax.com...
> >> -- AND RAID 0 will double the chance of a HD
> >> failure.
> >
> >Nope, the chance of the failure of a RAID 0 array with two drives is
about
> >double that of a single drive. In a RAID 0 array the chances of any
drive
> >failing remains unchanged.
>
> Inasmuch as a RAID 0 array does the job that a single disc would be
> doing otherwise, I think you just said what I said, didn't you??

Not exactly which is why I posted a very precise statement on the issue.

 

[milleron]

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

On Thu, 29 Jul 2004 08:13:44 +1000, Paul Taylor
<birder@ozemail.com.au> wrote:

>Milleron wrote:
>> Now, how in the world could RAID 1 give ANY increase in read
>> performance, let alone "double" the rate?? If we're talking about
>> theory, the latency in the RAID controller could only SLOW BOTH read
>> and write performance.
>
>Software RAID-1 under Solaris (i.e. DiskSuite) uses round-robin reads
>to the drives in the array; the read performance a RAID-1 on an MSI
>motherboard at work (Promise RAID controller, I think) was substantially
>better than with a single drive, so I assume it uses a similar arrangement.
>This only applies to reads (in particular sequential reads) - writes have
>to go to both drives, and would be slightly worse than a single drive due
>to the RAID overheads.

I stand corrected! Thanks very much for the explanation.
Ron