RE: Oracle Standard Edition & RAC

Using RAC for scalability is a popular misconception and a great MIPS (Marketing Invention for Pushing Sales). Synchronization between nodes takes a lot of resources. Communication over network interface is several orders of magnitude slower then IPC. Diagnostics is extremely hard and global enqueues can kill an OLTP application very quickly. Eventually, the complexity, wasted resources and exorbitant amounts of money spent on the needed equipment to provide illusion of HA will eventually generate a backlash against Oracle. Fortunately, Oracle has no real competition in the market, so that backlash will mostly be reduced to nagging, like the one in this post. Limited high availability, approximately equal to the one provided by cheap RAC configurations can usually be provided by NUMA technologies. What people don't understand is that Altix, Superdome or P595 can provide the same or higher level of uptime as clustered Dell boxes with much, much better and more predictable performance. Unfortunately, NUMA is not such a buzzword as RAC.

Mladen Gogala
Sr. Oracle DBA
1500 Broadway, 6th floor
New York, NY 10036
(212) 329-5201
www.vmsinfo.com

The Leader in Integrated Media Intelligence Solutions

> -----Original Message-----
> From: oracle-l-bounce_at_freelists.org [mailto:oracle-l-
> bounce_at_freelists.org] On Behalf Of Kevin Closson
> Sent: Friday, January 05, 2007 10:45 AM
> To: oracle-l_at_freelists.org
> Subject: RE: Oracle Standard Edition & RAC
>
>
> A single node 4 CPU SE should scale even better than 2 node x 2 CPU
> SE
> RAC so using SE RAC for scalability is luxury and company doing so
> has
> way too much money.
>
> ...very true
>
> Availability? Well, besides Oracle licensing, HA environment
> requires
> fair amount of investment if it's really HA and not just "reported"
> HA
> because of RAC.
>
> ...please explain?
>
> --
> http://www.freelists.org/webpage/oracle-l

--
http://www.freelists.org/webpage/oracle-l