Aye to that, but you'll need at least two, possibly
three, identical clusters, not just one. One cluster for production
and an identical cluster for QA/Test, and possibly one for development
(though that last is often regarded as unnecessary). Skimping on the
QA/Test environment is the leading edge of failure...
RAC itself requires additional DBA expertise as
well as additional OS SysAdmin expertise for cluster hardware/OS, each of which
costs more to obtain/maintain (either by hiring experienced/talented or
training to build or both). Clustering is not a low-cost solution
from any perspective...
RAC is a solution for certain specific
high-availability and high-scaleability requirements (not including
"data-center failure", a.k.a. disaster-recovery), so it's a good idea to be
certain that you are planning a solution that meets your own specific
requirements before proceeding. RAC should not be a high-level
management decision -- it is a specific technical solution to meet specific
technical requirements, which themselves should have been derived from the
requirements of the business. There are several other possible H/A
solutions in Oracle9i (i.e. physical standby, logical standby, advanced
replication, OS failover solutions, RAC, etc), each of which addresses the same
H/A problems in different ways with differing levels of complexity and
cost.
|
- Oracle Real Application Clusters prem
- RE: Oracle Real Application Clusters Jamadagni, Rajendra
- RE: Oracle Real Application Clusters Tim Gorman
- RE: Oracle Real Application Clusters Jamadagni, Rajendra
- RE: Oracle Real Application Clusters Loughmiller, Greg
- RE: Oracle Real Application Clusters DENNIS WILLIAMS
- Re: Oracle Real Application Clusters Tim Gorman
- RE: Oracle Real Application Clusters prem
- Re: Oracle Real Application Clusters prem
- Re: Oracle Real Application Clusters Ray Stell
- RE: Oracle Real Application Clusters Jamadagni, Rajendra
- RE: Oracle Real Application Clusters DENNIS WILLIAMS
- Re: Oracle Real Application Clusters Tim Gorman