Throughout this book we have discussed the concepts and architecture behind Oracle’s parallel execution and parallel server features. We also discussed various aspects of administering, monitoring, and tuning these features. This appendix takes the discussion of Oracle’s parallel support to its logical conclusion by talking about the types of applications that can benefit from parallelism.
The three case studies chosen for this appendix cover the three most common hardware architectures used for parallel systems: SMP, MPP, and cluster. The case studies also cover two common types of applications: online transaction processing (OLTP) systems and decision support system (DSS) applications. The case studies include:
An OLTP application on a Sun SMP platform
An OLTP application on an IBM RS/6000 HACMP cluster
A DSS application on an IBM SP, which is an MPP platform
Although these case studies discuss only Sun and IBM platforms, they are equally applicable to all platforms that support Oracle’s parallel processing features.
When you consider an application’s suitability for use with Oracle’s parallel feature set, you really need to consider two things separately. First, you need to look at your application with respect to Oracle Parallel Server (OPS). Not all applications are well-suited for OPS, and in some cases switching to OPS could drastically hurt performance and throughput. Second, you need to consider whether your application ...