Re: RAC High End Scalability?

"Stephan Bressler" <stephan.bressler_at_siemens.com> wrote in message news:b2tg11$771$1_at_news.mch.sbs.de...
> Hi Michael,
>
> setting up RAC successfully highly depends on whether or not the
application
> can be partitioned into logical units (e.g. sales europe/asia/north
america)
> or whether you run independent applications on the nodes (e.g. operational
> system and DWH).
> If the nodes access (modify!) the same dataset frequently, RAC is not an
> option without an database/application redesign.
>
> But if this is not possible, mssql isn't an option at all. This one
> literally relies on partitioning. The URL
> (http://www.microsoft.com/sql/evaluation/features/distpart.asp) mentioned
in
> the documents reads:
>
> The Distributed Partitioned Views feature shares the database-processing
> load across a group of servers,
> !!!! by horizontally partitioning the SQL Server 2000 data !!!!
>
> in the first sentence.
>
> It may be true that oracle does not scale well in some (a lot?) of
> scenarios, but I believe mssql to perform even worse.
>

The setup for sql server clustering seems to be the bigger problem to me.

I could be wrong, but doesn't the sql server clustering require actually splitting data into different copies for each node and then merging the data with a view ?

Who in their right mind would implement something like that ?

It seems misleading to even allow in tpc.org benchmarks something (like this) that no one would put in a real production database.

> BTW: TPC-C is partitioned horizontally by design.
>
> Stephan
>
>
> "Michael Harvey" <mgharvey_at_bigpond.com> wrote in message
> news:5b945v88h9rg40to39sab6cauio7r855gj_at_4ax.com...
> >
> > ORACLE make lots of claims about the scalability of RAC that I would
> > like to believe.
> >
> > Microsoft recently (late last year) published a paper disputing some
> > of these claims.
> >
> > http://www.microsoft.com/sql/evaluation/compare/RealityBehind_RAC.pdf
> >
>
>
>
Received on Tue Feb 18 2003 - 21:07:47 CST