Module 10
Monitoring and Optimizing a BI
Solution
Module Overview
• Overview of BI Monitoring
• Monitoring and Optimizing the Data Warehouse
• Monitoring and Optimizing Analysis Services
• Monitoring and Optimizing Reporting Services
Lesson 1: Overview of BI Monitoring
• Introduction to Monitoring and Optimization
• What to Monitor
• Creating a Performance Baseline
Introduction to Monitoring and Optimization
Business Requirements
Technical
Data
Architecture
Warehouse Reporting and
and
and ETL Analysis Design
Infrastructure
Design
Design
Monitoring and Optimizing
Operations and Maintenance
What to Monitor
• Health monitoring
• Periodically monitor hardware
resource utilization
• Identify problems early
• Performance troubleshooting
• Respond to perceived
performance issues
• Typically these problems are as a
result of changes in:
• Response time
• Throughput
Creating a Performance Baseline
• Collect data over a representative period of time
• Include normal workload cycles, such as:
• Data warehouse load operations
• Data model processing operations
• Normal everyday query activity
• Peak-time query activity
• Use Performance Monitor to create a data
collector set
• Collect general system metrics on CPU, memory,
disk and network usage
• Retain these to use as a baseline
Lesson 2: Monitoring and Optimizing the Data
Warehouse
• Considerations for Data Warehouse Performance
• Monitoring Tools for the Data Warehouse
• Using Resource Governor to Balance Resource
Utilization
• Troubleshooting Data Warehouse Performance
Considerations for Data Warehouse Performance
• Workloads:
• ETL data load
• Data model processing
• Report queries
• Operational activities
• SQL Server Settings
• Memory
• Processor
Monitoring Tools for the Data Warehouse
• Performance Monitor
• SQL Server Profiler
• Database Engine Tuning Advisor
• Data Collector
• Dynamic Management Views
Using Resource Governor to Balance Resource
Utilization
• Use Resource Governor to define maximum and
minimum thresholds for:
• CPU usage
• Memory usage
• IOPS
• To configure Resource Governor:
• Create resource pools with defined thresholds
• Create a workload group for each workload type and
associate it with a resource pool
• Create a classifier function to allocate sessions to
workload groups
Troubleshooting Data Warehouse Performance
• Use Performance Monitor, SQL Server Profiler,
and dynamic management view to troubleshoot
• When troubleshooting, a useful approach is to
review performance metrics in this order:
1. Processors
2. Memory
3. Storage
Lesson 3: Monitoring and Optimizing Analysis
Services
• Considerations for Analysis Services Performance
• Monitoring Tools for Analysis Services
• Troubleshooting Analysis Server Queries
Considerations for Analysis Services Performance
• Analysis Services must
support:
• Data model processing
workloads
• Users query workloads
• Analysis Services settings
for improving and
troubleshooting
performance include:
• Memory management settings
• Flight Recorder
Monitoring Tools for Analysis Services
Key tools for monitoring Analysis Services include:
• Performance Monitor
• SQL Server Profiler
• Dynamic Management Views
Troubleshooting Analysis Server Queries
Query Results
Session Manager
Session Initialize
Query Start Query Processor Query End
Query Cube End
Query Cube Start
Serialize Results End
Query Processor Cache
Serialize Results Begin
Query Subcube Start Storage Engine Query Subcube End
Data Cache
Get Data From Aggregation Get Data From Cache
Dimension and Measure Group Data
Lesson 4: Monitoring and Optimizing Reporting
Services
• Considerations for Reporting Services
Performance
• Monitoring Tools for Reporting Services
• Using Caching and Snapshots to Optimize
Performance
Considerations for Reporting Services
Performance
• Reporting Services
workloads
• Data retrieval and processing
• Report rendering
• Reporting Services settings
• Memory configuration
settings in the in the
[Link] file
Monitoring Tools for Reporting Services
• Performance Monitor is the primary tool for
monitoring Reporting Services
• You can use the following Performance Monitor
objects
• Web Service
• MSRS 2014 Web Service
• MSRS 2014 Web Service SharePoint Mode
• Windows Service
• MSRS 2014 Windows Service
• MSRS 2014 Windows Service SharePoint Mode
• HTTP and Memory
• ReportServer:Service
• ReportServerSharePoint:Service
Using Caching and Snapshots to Optimize
Performance
Cached Snapshot Report
Cached
Shared Report History
Dataset
ReportServerTempDB ReportServer
Lab: Monitoring and Optimizing a BI Solution
• Exercise 1: Monitoring and Optimizing a Data
Warehouse
• Exercise 2: Monitoring and Optimizing Analysis
Services
• Exercise 3: Monitoring and Optimizing Reporting
Services
Logon Information
Start 20467D-MIA-DC and 20467D-MIA-SQL, and
then log on to 20467D-MIA-SQL as
ADVENTUREWORKS\Student with the password
Pa$$w0rd.
Estimated Time: 75 Minutes
Lab Scenario
You have implemented a BI solution for Adventure
Works Cycles. Now you must monitor the
performance of the solution, and optimize it
where necessary. The solution consists of a data
warehouse that provides a source of data for a
multidimensional Analysis Services data model
and several Reporting Services reports.
Lab Review
• How might the classifier function you would
create to prioritize ETL workloads in a real solution
differ from the one used in the lab?
Module Review and Takeaways
• Review Question(s)
