ST230c: Performance Administration on Clustered Data ONTAP

Agenda:

The content is based on the current official NetApp™ course materials for the "ONTAP Performance Administration" course and is supplemented by the instructor's 27 years of ONTAP experience, corresponding additional commands, and even deeper interpretations and checklists.

• Building a NetApp™ Cluster
  • Operating System Layers
  • Data Flow in the Cluster
  • Troubleshooting Configuration Errors (Hardware and Software)
    
    
• Introduction to Challenges
  • Physical Limits and Additional Efforts for Error Prevention or Correction: Latency and Response Times, Fragmentation, Misalignment, Small IOs, Reconstructs, Silent Data Corruption
  • Fairness Brakes in NAS and SAN Protocols and ONTAP Management Layers: TCP Window Size, Buffer Credits, Quality of Service, CMD Queue Limits
  • How can we recognize them?
  • What is their impact?
    
    
• Introduction to NetApp™ Countermeasures
  • Lost Write Protection, DIF/DIX=PI, WAFL Write Optimization, RAID-DP/TEC, Free Space Reallocate, Read Reallocate, Read Ahead, Dedupe, Compression, Compaction, VAAI, ODX, Thin Provisioning, Quality of Service, Volume Move, FlexGroups, InodePath
    
    
• The Mathematical Foundations
  • Queueing Theory and its Concrete Application in Storage and Network Performance Analysis
  • Recognizing Open and Closed Load / Single-threaded and Parallel Workloads
  • Criteria for Identifying and Distinguishing Bullies and Victims
    
    
• Introduction to Commands for Identifying
  • Highest Latencies
  • Highest IO Loads
  • Highest Data Throughputs
  • Largest Queues
    
    
• Performance Tuning
  • HDD + PAM = Flash Cache
  • HDD + SSD = Flash Pool (Hybrid Aggregate)
  • All Flash FAS (AFF)
  • FlexGroup
  • Optimizations: Which are compatible? Which cancel each other out?
    
    
• Sizing
  • Adequate Consideration of Client IO Workloads
  • Estimation of Additional Housekeeping and Metadata IOs
  • Size of Metadata and Working Set
    
    
• Overview of Additional External Monitoring Tools
  • System Manager
  • Active IQ Unified Manager
  • Harvest, Graphite, Grafana, NABox
    
    
• All topics of the official NetApp™ training "ONTAP Performance Administration" are covered and deepened:
  • Describe how to use NetApp™ tools for performance measurement
  • Explain the benefits of using Active IQ for performance analysis
  • Use Active IQ Unified Manager to monitor the performance of cluster objects
  • Describe the layers of ONTAP software architecture
  • Diagram the flow of read and write requests through the ONTAP network and protocol layers
  • Use performance analysis tools to identify WAFL-related performance bottlenecks
  • Discuss how storage quality of service (QoS) operates in an ONTAP environment
  • Explain how to monitor and manage workload performance
  • Use performance analysis tools to identify NAS performance bottlenecks
  • Identify the actions that you can take to resolve SAN performance issues
    
    

Objectives:

Target audience:

Prerequisites: