LWN.net Logo

Oracle on Linux edges SQL Server in benchmark (Register)

The Register reports on the results of a recent Oracle benchmark, where Linux was 14 percent faster than Microsoft SQL Server. "The results come from running Oracle9i Database Release 2 with Real Application Clusters on Linux against Microsoft SQL Server 2000 on a 32-processor cluster configuration, with identical processors and the same amount of memory per CPU."
(Log in to post comments)

Nothing like an apples to truck tires comparison...

Posted Sep 18, 2002 13:20 UTC (Wed) by sphealey (guest, #1028) [Link]

What is the value of such numbers? Even to an evil Marketing department? My 10 year old child would immediately point out that benchmarking two applications on two different platforms introduces so many variables that the comparison is worse than useless. This isn't even an apples to oranges comparison: it is more like apples to floor wax.

sPh

Very poorly written

Posted Sep 18, 2002 14:52 UTC (Wed) by proski (subscriber, #104) [Link]

The same text "Oracle achieved 138,362.03 tpmC and a price performance of $17.21/tpmC" appears twice in the article. There is too few information about the hardware, and even less information about the software, on top of which Microsoft SQL server was running. Was it Windows NT, 2000, XP or maybe (just kidding) Wine on Linux?

Editors, please ignore poor articles. And be careful with The Register.

Oracle on Linux edges SQL Server in benchmark (Register)

Posted Sep 18, 2002 16:38 UTC (Wed) by lance (guest, #2263) [Link]


And how does PostgreSQL stack up?

Oracle on Linux edges SQL Server in benchmark (Register)

Posted Jun 24, 2003 11:48 UTC (Tue) by johanroos (guest, #12360) [Link]

The apples narrative is fine but what do you do if you can't compare SQL server to anything else because it only supports Microsoft?
So just because you can't compare the two on Linux means that SQL server will always be better because it is only for Microsoft?

This is a very valid comparison because it compares a SOLUTION and not a DB.
The bottom line is the time experienced by the user hitting you DB. That is what must be compared.

Copyright © 2002, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds