Windows Server 2003系统效能监视文稿.pptVIP

TechNet TNT1-16 Windows Server 2003系統效能監視 林寶森 jeffl@ms11.hinet.net Why Monitor Performance? Analyze performance data to uncover bottlenecks By monitoring performance, you obtain data that you can use to: Understand your workload and the corresponding effect on your systems resources Observe changes and trends in workloads and resource usage so you can plan for future upgrades Test configuration changes or other tuning efforts by monitoring the results Diagnose system problems and identify components or processes for optimization System Bottlenecks What Is Server Analysis and Optimization? What Is Task Manager? What Is the Performance Console? What Is Real-Time and Logged Monitoring? System Monitor Objects, Instances, and Counters Objects Are Major Components or Subsystems of the Computer System Instances Are Multiples of the Same Object Counters: Are Measurements of different aspects of objects Continually gather data on objects Provide data on all instances of an object Can be selected to displayed in System Monitor Viewing Counter Data Identify and Resolve Memory Bottlenecks Virtual Memory Management The Windows Memory Model Identify and Resolve Processor Bottlenecks Identify and Resolve Disk Bottlenecks Identify and Resolve Network Bottlenecks What Is a Counter Log? Each performance object provides performance counters that represent data about specific aspects of a sytem or server Counter logs define what data is stored in the log file Counter Log File Formats What Is an Alert? Feature that detects when a predefined counter value rises above or falls below a specified setting Specified setting on the counter is called alert threshold Set an alert on a counter when: Make an entry in application event log Start a predefined counter log Send a Message Run a Program Set alerts based on established performance baseline values Use alerts to be noti