IBM Cognos PowerCubes allowed organizations to gain valuable insights into many areas of the business, find trends, inter-dependencies and drive efficiency of business processes.
In order to derive full benefits of multi-dimensional analysis in Analysis, Report and Metric studios cubes are supposed to perform fast.
To get the best out of cubes we have developed PowerCubes Performance Assessment Methodology (PC PAM). Based on IBM Cognos best practices recommendations, supported with dozens of successful implementations and complimented by our consultants’ years of experience our methodology provides significant benefits to our clients:

Helps troubleshoot performance and design bottlenecks
Identify optimization options available for specific cube environment and design
Develop sound and effective cubes maintenance process
Most important – keep business users satisfied

To get the best out of cubes - PowerCubes Performance Assessment Methodology

Outcomes:

Reduced cubes build time
Simplified maintenance resulting in fewer hours and resources required
Streamlined new development
Better performance of reporting and analytical applications

Our Performance assessment methodology is designed to provide comprehensive and all-inclusive solutions for analysis and optimization of medium to large-scale BI environments.

It consists of:

• Model Analysis
• Cube Design Assesment
• Build Environment Evaluation and
• Maintenence Planning / Execution

We've also developed a program that goes far beyond technical aspects of BI optimization and involves all components of business cycle, including:

• Business Requirements Evaluation – what users want
• Usage Patterns: what user need - recognize, discover, identify (reports, analysis, dashboards, metrics, alerts)
• Model Analysis and Validation (data sources, dimensions, measures, custom members, calculations),
• Cube Design: Uniqueness, Slow Changing Dimension, Grain, Drill-Trough
• Processing Environment Estimates: Hardware/Memory/Resources
• Performance benchmarking: Before & After
• Health Card (help file) – really quick reference
• Overall System Optimization: $ how mach can we save?
• Recommendations for further improvements – establishing continuous cube maintenance process and routines

• Better run-time performance
• Reduced build time
• Simplified maintenance resulting in fewer hours and resources required
• Streamlined new development

Model Anallysis: Design considerations for slow performance in Analysis Studio tests:

• Data Sources Design and/or Optimization
• Cube Group
• Cube Partition
• Multi-file Cube
• Cross-tab caching
• Drill-Through
• Incremental Updates set up
• Model Elements (measures, categories, calculations,etc.)

Design Assessment: Cubes must be analyzed from reporting / analysis prospective:

• Report Calculations
• MDX – accessing leaf members
• Circular references
• Master – details
• Appropriate Drill – Through
• Report Usage

Build Environment: Considerations to reduce build time:

• Multiprocessing
• Memory Allocation
• Temporary files – sufficient space requirements
• Optimized I/O processing
• Optimized OS Environment
• Optimized build environment configuration
• Configuring database access settings
• Practical limits of model and cube size

Maintenance: The basic steps of cube maintenance plan should include:

• Source Data Updates
• Model Updates
• Cube Updates
• Multi-file Cubes
• Match Model and Source Columns
• Categories Move with source data changes
• Delete inactive categories
• Recover Failed Model
• Modifying Power Cubes
• Proper Partitioning Strategy
• Clean up the model periodically