Designing User Interfaces Spring 1999

1
SE 767-NT Software Performance Engineering Robert
Oshana Lecture 15 For more
information, please contact NTU Tape
Orders NTU Media Services [phone]
oshana_at_airmail.net
tapeorders_at_ntu.edu
2
Where we are

• Introduction
• SPE Quick View
• SPE and the UML
• Software Execution Models
• Web applications and other distributed systems
• System execution models
• SPE data collection
• Software measurement and instrumentation
• Performance oriented design
• Performance patterns
• Performance anti-patterns
• Implementation solutions
• Web applications
• Embedded and real-time systems
• The SPE process
• Implementing SPE

3
Resource estimation techniques

• Use measurements
• Study measurements
• Use a mentor
• Best-worst case estimates
• What to estimate
• Estimating I/O requirements

4
More one estimating I/O requirements

• Need two types of information
• Number of physical I/Os
• I/O service times
• Two ways to estimate physical I/Os
• Estimate logical I/Os and sensitivity to hit
  ratios and write ratios
• Directly estimate which I/O requests are likely
  to be found in or written from a buffer or a
  cache and only count physical I/Os

5
More one estimating I/O requirements

• A typical estimate
• Ip (i 1) X (1 h)
• Read requests that sequentially access a file,
  number of physical I/Os is n/b
• Some I/O devices read one track with each
  physical I/O
• Read request to a DB depend on complexity of SQL
  statement
• SELECT ltfieldsgt from table where keyx

6
Estimating network messages

• Estimate the number of messages transferred or
  the number of characters transferred
• Account for headers as well as application
  messages
• Mi d / (m h)
• Mo (r x s) / (m h)

7
Obtaining computer resource requirements

• Need computer resource requirements for each SW
  resource requirement estimated
• May be able to obtain estimates for computer
  resource requirements from performance
  specialists
• Make sure to account for subtle overhead factors
• Unpacking a message before using

8
Summary

• Need the following data for SPE
• Performance objectives
• Key performance scenarios
• Execution environment
• Software resource requirements
• Computer resource requirements
• Walkthroughs used to collect data

9
Software measurement and instrumentation

• Chapter 8

10
Where we are

• Introduction
• SPE Quick View
• SPE and the UML
• Software Execution Models
• Web applications and other distributed systems
• System execution models
• SPE data collection
• Software measurement and instrumentation
• Performance oriented design
• Performance patterns
• Performance anti-patterns
• Implementation solutions
• Web applications
• Embedded and real-time systems
• The SPE process
• Implementing SPE

11
Introduction

• Measurement is a key part of the SPE process
• Data for SPE models
• Verify and validate models
• Determine whether performance goals have actually
  been met
• Monitor system performance over its lifetime

12
Introduction

• Obtaining the right measurements is not easy
• Tools can help
• Also need a strategy

13
What should you measure?

• System understanding
• Model specifications
• Model updates
• Model verification and validation
• Software performance evaluation

14
Types of measurement data

• Workload data
• Data characteristics
• Execution characteristics
• Path characteristics
• Software resource usage
• Processing overhead

15
Types of measurement data

• Computer system usage
• Scenario response time
• Scenario throughput
• Key computer system resource usage
• Resource utilization
• Resource throughput
• Resource queue lengths

16
(No Transcript)
17
Planning for performance measurement

• Key considerations
• Representative
• Reproducible
• Workload selection
• Load drivers
• Software
• Computer system environment

18
Performance benchmarks

• Hardware or software system evaluations
• Volume testing
• Performance measurement
• SPEC
• EEMBC
• BDTI

19
SE 767-NT Software Performance Engineering Robert
Oshana End of lecture For
more information, please contact NTU Tape
Orders NTU Media Services [phone]
oshana_at_airmail.net
tapeorders_at_ntu.edu