I would have to agree with your assesment. I would definately go with Linux
because in my experience, it seems to never crash. In fact, I don't ever
recall a time that Linux just up and crashed! Xfree86 froze, but all I did
was press Ctrl-Alt-backspace. I chose Linux even when I thought NT was
faster. Not only does Linux seem to run well, but I enjoy working with it!
:)
> -----Original Message-----
> From: jonathan [mailto:jonathan]On Behalf Of Jonathan Kovacs
> Sent: January 16, 2000 9:31 PM
> To: Trenton D. Adams; [EMAIL PROTECTED]
> Subject: Re: Linux & Samba vs NT
>
>
> "Trenton D. Adams" wrote:
> >
> > A while back I heard that NT outperformed Linux in networking
> in a message
> > from this list! The person writing the message even gave a link to a
> > Microsoft site that stated this FaCt??? I was very silly, and simply
> > trusted microsoft not to be biased. I don't know what I was
> thinking!!! I
> > have now found information that states that when tested with Ziff Davis
> > NetBench, Linux and Samba outperform NT. Not only that, but as
> we all know,
> > Samba integrates into an NT domain seamlessly!
>
> > NetBench, Linux and Samba outperform NT. Not only that, but as
> we all know,
> > Samba integrates into an NT domain seamlessly!
> >
> > http://lwn.net/1999/0121/samba.html
>
> Ah yes, the ever so painful ZD / Mindcraft fiasco. IIRC, the way it
> went was that ZD tested Linux+Samba vs NT and Linux 'won'. Then, an
> 'independant' company, Mindcraft tested NT vs Linux as a server and
> found that NT was superior (The test was paid for by Microsoft). Why?
> Because given the right test parameters any benchmark can be made to say
> anything you want. Benchmarks tend to have little relevance to
> real-world performance. I mean, sure Server A may be 30% faster than
> Server B, but if that comes at the cost of significantly higher downtime
> / management costs, then Server A isn't really that good after all.
>
> In the real world, it's been my experience that on low-end to moderate
> hardware, Linux will mop the floor with NT. On high end machines with
> large RAID arrays, >2 Gig of RAM and 4+ >=100Mbps NICs, NT tends to
> scale better. (Although, it's my personal opinion that when you're
> looking at that kind of hardware, you're better off with a SPARC, but
> that's just me...but I'm a UNIX bigot.) I hear that the main goal of the
> 2.4 kernel is to improve scalability, so this may change in a few
> months.
>
> Ultimately, it's probably a better idea to look at the resources you
> have and your project requirements than worry over perceived performance
> differences. Just keep in mind that no matter what you put into place,
> it's not going to last forever and will have to be upgraded at some
> point anyways, so it's better to go for a system that will be easy to
> maintain and will fit your specific requirements, than with a system
> that ZD/Mindcraft/<insert benchmark here> says is x% faster *under test
> conditions*.
>
> Jonathan Kovacs
>
-
To unsubscribe from this list: send the line "unsubscribe linux-net" in
the body of a message to [EMAIL PROTECTED]
