Schulung - IBM CV964G - Db2 12 for z/OS SQL Performance and Tuning

Course Outline

Introduction to SQL performance and tuningPerformance issuesSimple exampleVisualizing the problemSummaryPerformance analysis toolsComponents of response timeTime estimates with VQUBE3SQL EXPLAINThe accounting traceThe bubble chartPerformance thresholdsIndex basicsIndexesIndex structureEstimating index I/OsClustering indexIndex page splitsAccess pathsClassificationMatching versus ScreeningVariationsHash accessPrefetchCaveatMore on indexesInclude indexIndex on expressionRandom indexPartitioned and partitioning, NPSI and DPSIPage range screeningFeatures and limitationsTuning methodology and index costMethodologyIndex cost: Disk spaceIndex cost: MaintenanceUtilities and indexesModifying and creating indexesAvoiding sortsIndex designApproachDesigning indexesAdvanced access pathsPrefetchList prefetchMultiple index accessRuntime adaptive indexMultiple table accessJoin methodsJoin typesDesigning indexes for joinsPredicting table orderSubqueriesCorrelated subqueriesNon-correlated subqueriesORDER BY and FETCH FIRST with subqueriesGlobal query optimizationVirtual tablesExplain for subqueriesSet operations (optional)UNION, EXCEPT, and INTERSECTRulesMore about the set operatorsUNION ALL performance improvementsTable design (optional)Number of tablesClustering sequenceDenormalizationMaterialized query tables (MQTs)Temporal tablesArchive enabled tablesWorking with the optimizerIndexable versus non-indexable predicatesBoolean versus non-Boolean predicatesStage 1 versus stage 2Filter factorsHelping the optimizerPaginationLocking issuesThe ACID testReasons for serializationSerialization mechanismsTransaction lockingLock promotion, escalation, and avoidanceMore locking issues (optional)Skip locked dataCurrently committed dataOptimistic lockingHot spotsApplication designAnalyzing lock waitsMassive batch (optional)Batch performance issuesBuffer pool operationsImproving performanceBenefit analysisMassive deletes

Objective

After completing this course, students will be able to:

• Understand and design better indexes
• Determine how to work with the optimizer (avoid pitfalls, provide guidence)
• Optimize multi-table access
• Work with subqueries
• Avoid locking problems
• Use accounting traces and other tools to locate performance problems in existing SQL
• and more