[Gllug] Recommended distro

Christian Smith csmith at micromuse.com
Mon Dec 13 17:14:06 UTC 2004


On Mon, 13 Dec 2004, Jack Bertram wrote:

>* Joel Bernstein <joel at fysh.org> [041213 11:46]:
>> > He's probably made the "common mistake" of actually benchmarking a Sun
>> > machine against a comparably priced Intel box and finding out that the
>> > Sun machine is around 3 times slower.
>>
>> Perhaps so, but CPU MHz isn't the best indicator of computational power
>> between CPUs of similar architecture, let alone totally different ones.
>> I wasn't commenting on pricing of hardware etc.
>
>But speed on its own is fairly meaningless as a metric -
>price/performance is the most important comparator for nearly any use
>which is speed dependent (other applications may put more emphasis on
>reliability/price or capacity/price or some other tradeoff).  Which is
>why Martin originally put them together - too expensive and too slow.


In a commercial environment, I'd have thought reliability and support are
major contributors to whether to choose PCs or SUN kit.

For CPU clusters, yes, go for the cheap, throw away PCs. Reliability is
not of as great concern as performance.

But for single point of failure servers? If fast enough is fast enough, go
for the most reliable solution. Price/performance means nothing if support
costs of failed hardware dwarf initial purchase cost (likely). When a
whole department relies on a single server, that price/performance will be
for nothing if the whole department is twiddling it's thumbs waiting for
it to be fixed all afternoon.


>
>j
>

Christian

-- 
    /"\
    \ /    ASCII RIBBON CAMPAIGN - AGAINST HTML MAIL
     X                           - AGAINST MS ATTACHMENTS
    / \
-- 
Gllug mailing list  -  Gllug at gllug.org.uk
http://lists.gllug.org.uk/mailman/listinfo/gllug




More information about the GLLUG mailing list