Document Outline

• HP Capacity Advisor 7.2 User Guide
• Contents
• 1 Introduction
  • Setup considerations
    • Licensing requirements
    • Installation
    • Collection methods
    • Disk space requirements
    • Credentials
    • Dependencies
  • Upgrades and reinstallation
    • Upgrading agents on HP-UX or OpenVMS managed systems
  • Navigating within Matrix OE and Capacity Advisor
• 2 Features
  • Data collection
    • Data collection infrastructure
    • Supported collection configurations
    • Comparison of agentless and UP data collection
    • HP ICperf data differs from Utilization Provider data
      • Differences in memory data
      • Differences in network data
        • Related topics
  • Experimentation
  • Sizing for service level objectives
  • Modeling considerations
  • Capacity Advisor commands
• 3 Key Capacity Advisor concepts
  • Capacity planning goals
    • Quality of service
  • Utilization monitor, calculator, and simulator
    • Data handling for virtual machines
    • Data normalization
  • Resources for which Capacity Advisor collects data
    • Upper bounds
  • Measuring and analyzing resource utilization
    • Peaks and sums
    • Sampling interval
    • Headroom
      • Headroom star rating
        • Interpreting the headroom star rating
          • Interpreting the star rating given by the HP Smart Solver
    • Missing or invalid data
      • The affect of invalid data on HP Smart Solver solution finding
    • Utilization limits
      • Specifying utilization limits
        • Sustained time limits
        • Percentage of time limits
      • Understanding utilization limit messages
        • Percentage of allocation
        • With sustained limits
        • With percentage of time limits
      • Scope of utilization limits
  • Adjusting power
    • Power cap data
  • Automating solution generation: HP Smart Solver
    • HP Smart Solver: Types of solutions
  • Determining trends in Capacity Advisor
    • Aggregation of points in business interval bins
      • Choosing an appropriate business interval
    • Exclusion of data points
      • Factors that affect data validity
    • Linear regression
    • Error analysis
  • Forecasting growth
    • The forecast model hierarchy
    • Forecast model attributes
• 4 Procedures
  • Accessing Capacity Advisor
  • Gathering data for Capacity Advisor
    • Data retention period
    • Impact of data collection on managed system performance
    • Data collection options
    • About the HP Capacity Advisor Data Collection Service
      • Solving utilization data collection problems
    • Configuring data metrics to better reflect your environment
    • Collecting data without using an agent
      • Listing systems currently in the configuration file
      • Setting advanced options in the agentless data collection file
    • Data collection and the HP Capacity Advisor Consolidation software
      • Using Capacity Advisor with consolidation software server data
      • Viewing the licensed servers
      • Working with Capacity Analysis
    • Importing data for use in Capacity Advisor
      • From the Optimize menu — Import HP OVPA Data
      • From the Optimize menu — Import HP ICperf Data
      • From the command line
      • Ascertaining the data collection availability for a set of servers
    • Possible sources of discrepancies in collected data
      • No data collected for a managed node
      • Data timestamp appears to be incorrect
      • Dynamic memory in HP Integrity virtual machines
  • Producing graphs and reports
    • Using the Profile Viewer
    • Using the report wizard
      • The report wizard
      • Creating an historic utilization report
      • Creating a scenario utilization report
      • Using the report wizard to create a scenario comparison
        • An example scenario comparison report
        • Calculating a virtualization consolidation ratio
      • Creating a consolidation candidates report
      • Creating a cost allocation report
      • Creating a peak summary report
      • Creating a population report
      • Creating a trend report
      • Creating a power report
  • Setting utilization limits
    • Setting global utilization limits
      • Adding a utilization limit
      • Removing a utilization limit
    • Setting workload utilization limits
      • Adding a utilization limit
      • Enabling or disabling the limit
      • Removing a utilization limit
    • Setting scenario-wide utilization limits
      • Adding a utilization limit
      • Enabling or disabling the limit
      • Removing a utilization limit
    • Setting scenario workload utilization limits
      • Adding a utilization limit
      • Enabling or disabling the limit
      • Removing a utilization limit
  • Forecasting utilization
    • Defining forecast models
      • Accessing the Global Forecast Model
        • Defining the Global Forecast Model
      • Accessing the forecast model for a workload or system
        • Defining the forecast model for a workload or system
      • Accessing the forecast model for a scenario
      • Accessing the forecast model for a workload within a scenario
      • Defining a forecast model
      • Disabling a forecast model
      • Enabling a forecast model
    • Generating forecasts
      • Viewing forecast data in a profile viewer
      • Viewing forecast data in a utilization report
  • Working with scenarios
    • Creating a planning scenario
    • Editing a scenario
    • Controlling the data display
      • Change the meter style selection
      • Change the data collection period by setting the data range
      • Change the meter representation
    • Copying a scenario
    • Renaming a scenario
    • Undo/edit/review applied what-if actions in a scenario
      • To view applied what-if actions
      • To edit applied what-if actions
        • Modifying resource utilization values
    • Deleting a scenario
  • Working with systems
    • Creating a system
    • Adding an existing system
    • Editing a system
    • Changing VMs to be servers
      • Change VMs to server
    • Changing servers to be VMs (manually)
      • Change servers to be VMs
    • Assigning VM hosts to a VMware DRS cluster in a scenario
      • Initially converting VM hosts to simulate a VMware DRS cluster
      • Adding VM hosts to a converted cluster
      • Removing a VM host from a VMware DRS cluster
      • Removing a VMware DRS cluster from a scenario
    • Moving a virtual machine
    • Removing a system
    • Editing network and disk I/O capacity
      • Getting there
      • Setting the upper bound values
  • Working with workloads
    • Introduction
    • Creating a workload
    • Editing a workload
    • Moving a workload
      • Move considerations
    • Parking a workload
    • Deleting a workload
  • Working with power
    • Calibrating power on actual systems
      • Calibrating a single system
      • Calibrating multiple systems at once
    • Calibrating power within a scenario
      • Getting there
      • Calibrating a single system
      • Calibrating multiple systems at once
  • Automating time-consuming simulations
    • Automated solution finding: System consolidation to VMs
      • Begin with a scenario
      • Choose the systems to consolidate
      • Step 1 of 3: Define the destination system(s) and attributes
      • Step 2 of 3: Set level of effort for Smart Solver calculation
      • Step 3 of 3: View Smart Solver solutions
        • Results: Automated consolidation to VMs
          • Expected results
          • Possible anomalies in the results
    • Automated solution finding: Load balance of servers or VM hosts
      • Begin with a scenario
      • Choose the VM hosts or servers to load balance
      • Step 1 of 3: Define the solution constraints
      • Step 2 of 3: Set level of effort for Smart Solver calculation
      • Step 3 of 3: View Smart Solver solutions
        • Results: Automated load balancing of servers or VM hosts
          • Expected results
          • Possible anomalies in the results
    • Automated solution finding: Workload stacking
      • Begin with a scenario
      • Choose the workloads to stack
      • Step 1 of 3: Stack the specified workloads onto servers
      • Step 2 of 3: Set level of effort for Smart Solver calculation
      • Step 3 of 3: View Smart Solver solutions
        • Results: Automated workload stacking
          • Expected results
          • Possible anomalies in the results
    • Getting more detail
• 5 Using Capacity Analysis
  • Introduction
  • Accessing Capacity Analysis
  • Exploring the Analysis tab
    • Readiness information
    • Working with the Analysis tab
      • Analysis tab menu bar
      • Server and data range selector
      • Analysis queries table
        • Filtering the query list
        • Customizing columns
      • Query history graph
      • Analysis query results table
      • Determining the time of the last analysis calculation
      • Printing the contents of the Analysis tab
  • Data collection for analysis queries
  • Predefined analysis queries
    • Finding underused systems
    • Finding overused systems
    • Controlling power costs
    • Finding systems with configuration issues
  • Creating custom analysis queries
    • Creating the query
      • Character support in Capacity Analysis queries
    • Constructing a filter
      • Filter creation tips
      • Using sub-expressions
      • When two expressions are false
    • Attribute selection
      • System Tab
      • CPU Tab
      • Memory Tab
      • Network Tab
      • Disk Tab
      • Power Tab
  • Custom query examples
    • Example: Finding large underused VMs that could be replaced with medium or small VMs
      • Background
      • Requirements
      • Defining the query
        • Creating and naming the query
        • Selecting the data to include in the query results
        • Defining the filter
      • Revising the query
    • Example: Reducing licensing costs by finding VMs with one too many cores
      • Background
      • Requirements
      • Defining the query
        • Creating and naming the query
        • Selecting the data to include in the query results
        • Defining the filter
  • Changing analysis queries
    • Editing an existing analysis query
    • Deleting an analysis query
• 6 Planning with Capacity Advisor
  • Getting ready
  • Task: Understand current resource usage
  • Task: Plan server consolidation
    • Understanding the consolidation task
      • Example consolidation: Stacking applications on an existing server
        • Step 1: Determine which systems to consolidate
        • Step 2: Create a scenario
        • Step 3: Edit the scenario
          • Adding a VM host to the scenario
          • Making servers become virtual machines
            • Adding processor cores
            • Continuation: Making servers become VMs
            • Changing the utilization limit for the scenario
    • Automating the consolidation task
      • Example consolidation: Automating stacking on a “what-if” server
        • Step 1: Determine which systems to consolidate
        • Step 2: Create a scenario
        • Step 3: Edit the scenario
          • Making servers become virtual machines using automated solution finding
            • Screen 1 of 3
            • Screen 2 of 2
            • Screen 3 of 3
  • Task: Estimating the effect of adding or moving processors
  • Task: Determining where to put a workload using automated solution finding
    • Example: Determining where to put a workload
  • Task: Estimating potential power savings across a range of server models
    • Example: Estimating potential power savings
      • Collect data for a period of time that fully reflects your business cycle(s).
      • Check power calibration on existing servers
      • Create a planning scenario
      • Create a “what-if” representation of a new HP server
      • Run a utilization report that includes power selections
      • Move workloads from the older servers to the newer server
      • Run a utilization report that includes power selections
  • Data modeling in the Scenario Editor
    • Data modeling for VSP vPars and VMs
• 7 Using Capacity Advisor with HP Serviceguard
  • Using Serviceguard to migrate Integrity VM guests
  • Data collection in a Serviceguard environment
• 8 Troubleshooting in Capacity Advisor
  • When there is no connection
  • Data is not being collected
  • Data may appear to be old when it is not
  • Data seems to be incorrect or lost
  • Data cannot be imported
  • Data analysis is not working in Capacity Analysis
  • Out-of-memory errors in the Java heap
  • Working with reports
  • Working in the scenario editor
  • Working with power
  • Workload issues
    • Workload and system relationships may not match when comparing Capacity Advisor and Matrix OE visualization information
    • Sub-OS workloads on Linux systems are not supported
  • Miscellaneous user interface issues
• 9 Capacity Advisor error and warning messages
  • Command error messages
  • Command warning messages
  • HP Smart Solver error messages
    • When available systems lack sufficient capacity
    • When the threshold for invalid data is exceeded
  • HP Smart Solver informational messages
• 10 Support and other resources
  • Information to collect before contacting HP
  • How to contact HP
  • Security bulletin and alert policy for non-HP owned software components
    • Subscription service
  • Registering for software technical support and update service
  • How to use your software technical support and update
  • HP authorized resellers
  • Related information
    • Related technical papers
• Calculation assistance
  • Cost calculation
    • Cost per kilowatt-hour
  • Cooling calculation
    • Cooling multiplier
  • Adjusting for platform changes
    • Memory multiplier
  • Adjusting for change in a workload
    • CPU workload multiplier
    • Memory workload multiplier
    • Network I/O workload multiplier
    • Disk I/O workload multiplier
    • Determining estimated utilization assumptions for a workload
  • Adjusting for virtualization changes
    • CPU Virtualization Overhead %
      • CPU Virtualization Overhead %
    • Hypervisor memory overhead
      • Hypervisor memory overhead
      • Doing the math for hypervisor memory overhead
        • HP Virtual Machine
        • VMware ESX 3
        • VMware vSphere
        • Microsoft Hyper-V
  • Power calibration calculations
    • Determining idle/max values for non-Blade servers
    • Determining idle/max values for Blade servers
      • P-class HP Blades and HP Blade system sizer
      • C-class HP Blades and HP Blade system sizer:
• Command reference
  • capcustombenchmark(1M)
  • capcustombenchmark (4)
  • capovpaextract (1M)
  • capprofile (1M)
  • capprofile (4)
  • capagentlesscfg
  • cappmpextract
• Units and terminology
• Snapshots of Capacity Advisor reports
  • Consolidation candidates report
  • Scenario comparison report
• Graphed data in Capacity Advisor reports
• Known problems in Capacity Advisor documentation
  • Errors in Capacity Advisor online help and command reference pages
• Documentation feedback
• Glossary
• Index