Re: DB2 Crushes Oracle RAC on TPC-C benchmark

> I was not talking about the cost of Oracle
> > licenses.
>
> In other words you were trying to change the subject. No one has ever
> tried to do that before.
> --
> Daniel A. Morgan

The TPC benchmarks between DB2 and Oracle use vastly different hardware architectures.

The Oracle benchmark uses 16 Intel Linux servers running 4 processors on each, for a total of 64 processors.

The DB2 benchmark uses a single UNIX server with 64 processors.

The raw performance of the DB2 config is about 2.5 times more than the Oracle (transactions per minute) for TPC-C. But the cost per transaction per minute is about the same (DB2 is slightly better).

But as you mentioned in your other posts, performance is not everything (where you trying to change the subject?). Cost of ongoing ownership is also important. You cannot tell me with a straight face that the ongoing network administration cost, UNIX/Linux administration costs, and DB administration costs of running 16 servers is the same as running one server. To make that claim is irresponsible.

I think that if you check with various outsourcing companies, they may not all charge $1,000 per month, per server (like the one I know about), but the cost of managing the 16 Oracle servers is going to be a lot higher than the cost of the 1 DB2 server. Unfortunately, most outsourcing agreements are for 5-10 years and corporations cannot easily change vendors.

Now clearly, there are some advantages to the Oracle RAC configuration that DB2 does not have, but cost of initial investment or cost of ownership is not one of them. Received on Sun Jan 30 2005 - 17:07:47 CST