Ok now that my name has been called out... I guess I will add my 3 cents worth 1) The decision to go RAC is a BIG decision monetarily, time, skill If your applications aren't mission critical why would you bother? If you can get away with a db and standby --- On Tue, 5/18/10, Yong Huang <yong321@xxxxxxxxx> wrote: From: Yong Huang <yong321@xxxxxxxxx> Subject: Re: Documentation for reasons to NOT use RAC? To: "Tony Adolph" <tony.adolph.dba@xxxxxxxxx> Cc: oracle-l@xxxxxxxxxxxxx Date: Tuesday, May 18, 2010, 10:14 PM > > I don't see how developers should write code differently >> if the code needs to be deployed to RAC.: > > Its not just being TAF/FAN aware, see Gints Plivna's comments > earlier in this thread. Tony, I understand Gints Plivna's and kathy duret's responses at //www.freelists.org/post/oracle-l/Documentation-for-reasons-to-NOT-use-RAC,23 They both proposed great ideas for application server admins. The only thing developers should do a little differently is perhaps "Use of parallel in queries to take advantage of the Rac power", but that's not really RAC-specific. In short, the developers don't need to do anything special for RAC. DBAs and mid-tier administrators do. Yong Huang -- //www.freelists.org/webpage/oracle-l