Re: Larry Ellison comments on Microsoft's benchmark

Richard Waymire wrote:

> The data is partitioned across each node for key tables. If a node fails
> any queries against the distributed partitioned view will fail (but NEVER
> return incorrect results). Hence the recommendation to run each node in an
> MSCS failover cluster.
>

Ok you mean the statement of L.Ellison is true without the incorrect results, right?

>
> Is shared-nothing clustering good for general systems? Ask just about every
> VERY large system in a cluster (Tandem, DB2, etc.).
>

My question was not "is sharing-nothing good" but "is THIS approach good for a general system. OK?
So think about scaling up or even scaling down of THIS system!! Let's suppose you want to add a 13th box to increase performance. Can you do it ? As the number of moths is always 12 you must use another way to partition the db -> you must rewrite whole app. Is it ok for you? Let's say you want to use this benchmark to implement your own system but you want to use 5 nodes (you just don't need 12). You'll have to use probably rather 6 instead of 5 (2 months per node) and you'll still have to rewrite your app....

From this point of view, i think the opinion of L.Ellison is ok.

>
> For an objective opinion on such matters, please read some relevant material
> such as "In Search of Clusters" by Pfister from IBM Corp. You might also
> look up some slides, etc. from Doctor Jim Gray
> (http://research.microsoft.com/~gray/). Before you dismiss the site because
> it's on Microsoft's web page, look at this credentials (including the Turing
> award).
>
Received on Fri Jun 30 2000 - 00:00:00 CDT