Chapter 2: RDBMS Features for Availability

---

Overview

Relax and get comfortable; the geeked-out turbo-powered propeller-head stuff is waiting for you later in this book. You know the goods we speak of-the tech so heady we felt compelled to put it in the book title: RAC, Data Guard, Flashback. RMAN, streams, and transparent application failover-the stuff to take your database to the next level.

But we have to cover a few of the less buzzy-sounding items first. Fact of the matter is, availability starts a long time before you get your Data Guard-protected 16-node flashback-enabled Oracle database up and running. Availability starts in the guts of your data architecture, inside the tables and indices and constraints and packages. The skeleton on which the meat of your data hangs must be built to withstand a nonstop, round-the-clock barrage of DML and query.

This chapter is dedicated to those little features inside the Oracle RDBMS that provide the groundwork for a high-availability (HA) environment. These are the features that drive your HA on a day-to-day basis, from the first day you go live. These are the features that allow you to change and morph on the fly, instead of taking massive outages to compensate for growth and (let's just admit it) shrinkage. You should note that all the internal features of the database that allow for high availability cannot possibly be covered in one chapter. What we have attempted here is a cross between a highlight reel and the standard college survey course: we hit the sexy features and provide a little depth to get your feet wet with each of them. Then, you can take what features you need to explore further and use our bibliography to research each of them further.

---