A properly configured parallel sysplex can deliver near continuous availability. "Properly configured" is often mistakenly equated with redundancy. But installations with highly redundant configurations have suffered significant outages. Why wasn't redundancy enough? What went wrong? In part 1, the speaker shares concepts and principles gleaned from analyzing customer installations under the auspices of IBM's High Availability Center of Competency (the HACoC team identifies risks to HA and suggests solutions). Part 2 provides practical application of these principles with suggestions for configuring and operating your sysplex in a way that maximizes the likelihood of achieving your availability goals for your critical business services.