Re: Microsoft destroys TPC-C records!

Norris (and other Windows folks),

I don't want you to take this the wrong way because I am not a Windows hater by any means (use quite a bit of 95/98 and NT myself in my work), but I think what we should be discussing here is the root of the problem with Microsoft's announcement.

I just can't see who is going to run a setup like the one used by Microsoft in this benchmark. We are talking about a multi-node (in this case 12) setup with a separate install of SQL Server on each node and a database split across each installation with custom distributed views to pull the needed information from all the scattered parts. It works like a bat out of hell as long as all parts are active, but one node hiccups (and come on, we are talking NT here. I like it but even 2000 gives the blue screen of death at will) and this database is useless.

There is no built-in redundancy in this case and the availability decreases as the number of nodes increases. Frankly, Microsoft is trying to redefine a common IT term again to suit their needs (in this case it's clustering).

No one running a 24x7 setup would ever go for something like this.

And again, it was pitted against a single Solaris box running a single 8i instance.

SQL Server has its place, but the VLB/high-availability arena is just not it.

Just my $0.02.

Later,
Gene....

Norris wrote in message <8ahhqj$1365$1_at_adenine.netfront.net>...
>Do you mean the fast performance of Windows 2000/SQLServer 8.0 is not
related to the speed of COM+ ?
>
>In comp.databases.sybase Michael D. Long <lead_dog_at_bellsouth.net> wrote:
>
>> I followed the link that you posted and now understand why this
>> middleware publication is confused about the part that COM+
>> serves in the TPC-C results.
>
>> *** If you look at the source code in Appendix A of the full disclosure
>> report
>> you will note that none of the features indicated above are even
referenced,
>> much less actually used!!!
>
>> I'm not trying to bash Microsoft, but I think misleading statements by
>> marketing
>> organizations are a disservice to all of us.
>
>> Mike
>
>> Norris wrote in message <8aa53e$2874$1_at_adenine.netfront.net>...
>>>To see MS innovations and how SQL Server 2000 achieved world record TPC-C
>> performance and price/performance results:
>>>
>>>http://msdn.microsoft.com/isapi/msdnlib.idc?theURL=/library/backgrnd/html
/m
>> egasrvs.htm
>>>
>>>
>>>In comp.databases.sybase Nuno Souto <nsouto_at_nsw.bigpond.net.au.nospam>
>> wrote:
>>>> On 9 Mar 2000 02:00:23 GMT, Norris <jcheong_at_cooper.com.hk> wrote:
>>>
>>>>>>> > Everything any culture on the planet has today is
>>>>>>> > ALL built on somebody else's work.
>>>
>>>> Don't have the slightest problem with this. That's what makes progress
>>>> work.
>>>
>>>> PROVIDED we are talking about "building on".
>>>
>>>> Badge-engineering and term re-invention like MS does all the time
>>>> is what I'm against. This is not the same as "building on somebody
>>>> else's work". This is crap, desguised as marketing.
>>>
>>>> Many years ago, it was the norm to use tar and feathers for what MS
>>>> marketing does. Nowadays, the stock market hikes up the share price
>>>> when MS re-names something instead of "building on something"...
>>>
>>>
>>>
>>>> Cheers
>>>> Nuno Souto
>>>> nsouto_at_nsw.bigpond.net.au.nospam
>>>> http://www.users.bigpond.net.au/the_Den/index.html
>>>
>>>
>>>--
>>>http://www.cooper.com.hk
>>>http://sybooks.sybase.com/onlinebooks
>
>
>
>
>--
>http://www.cooper.com.hk
>http://sybooks.sybase.com/onlinebooks
>
Received on Mon Mar 13 2000 - 11:30:24 CST