Agenda Items

• Performance Testing
 Performance Testing
Ensure application performs well under expected workload

• Application functionality is not the only concern

• Response time, reliability & resource utilization do matter

• Performance testing is not to eliminate bugs but bottlenecks

• Focus is on:
• Speed  determines application response time
• Scalability  determines maximum workload application can handle
• Stability  determines if application is stable under varying loads
 Why Performance Testing?
• Share application’s performance attribute with stakeholder (e.g. client)

• Uncover areas of improvement before application launch

• Poor performing application will lose customers, hence business

• It’s mandatory to perform performance testing for mission critical systems

• Space missions,
• Medical equipment with software

• Help determine life of system without malfunctions

 Types
Volume Testing
Load Testing
Stress Testing

Soak/Endurance Testing Spike Testing

 Load Testing
Applications ability to perform under anticipated user load

• Helps identify bottlenecks before system goes live or prior to busy season

• Constantly increase load until it reaches to defined threshold

• Increase in load  increase in number of concurrent users

• Determine response time, throughput, server condition & much more…

• Generally performed under controlled environment/situation

• Usually performed for important business transactions (e.g. ordering)

 Load Testing - Cases

• eCommerce website during FLAT 50% OFF

• FBRs e-filling portal during month of Sept

• Non-Stop transfer of files to & from a hard disk

• Download large file from a website

• Play same video by concurrent users

 Load Testing - Tools

• Numerous tools available

• Load Runner
• Blaze Meter
• Apica
• WebLoad

• How it works?
• For web based applications, increase load slowly until reaches to threshold
• Use multiple geographical locations
• Observe number of loops completed in a given time
• Monitor CPU/memory & database performance
• Report results
 Stress Testing
Meant to measure performance outside system parameters

• Tested with number of users or transactions beyond threshold limit

• Goal is to measure application stability

• Failure at what point system fails?

• Recovery  how does system recovers from failure?

• Determines how application behaves under unfavorable conditions (e.g. DDOS)

• Generally performed under controlled environment/situation

 Stress Testing – Why?

• It allows the test team to monitor system performance during failures.

• To verify if the system has saved data before crashing or NOT.

• To verify if the system prints meaningful error messages while crashing or random
exceptions.

• To verify if unexpected failures do not cause security issues.

 Stress Testing – How?

• Running several resource-intensive applications in a single computer at the same

time

• Attempt to hack into a computer and use it as a zombie to spread spam

• Flooding a server with useless e-mail messages

• Making numerous, concurrent attempts to access a single Web site or operation

• Attempting to infect a system with viruses, Trojans, spyware or other malware.

 Volume Testing
Test application with large amount of data
• Also referred to as flood testing

• Analyze system performance by increasing volume of data in DB

• Identify problems that are likely to occur with large volume of data

• Figure out point at which application degrades

• Help identify system capacity

 Volume Testing – What?

• To check if there is any data loss

• Check the system's response time

• Check if the data is stored correctly?

• Verify if the data is overwritten without any notification

• Check for warning and error messages, whether it comes at all for volume
problems

• Check whether high volume data affects the speed of processing

 Volume Testing – Recommendations
• Stop all servers and check all logs

• Before the load test, manually execute the application scenario

• For useful results, stagger the number of users

• To overcome license constraints, balance think time

• Be cautious with the new build

• Analyze the use case for improvement once a baseline has been established

• A repetition of particular parts of volume testing becomes inevitable in case there

is a performance bottleneck
 Soak Testing
Check the system performance for specified period of time

• Also referred to as “Endurance Test”

• System stayed under normal work load

• Executed or monitored for longer period of time

• Example:
• Suppose an OPD application is expected to run for 4hours a day; under soak
test, application will be executed for 8hours.

• Help determine issue like memory leaks or unexpected behavior

 Spike Testing
• It’s a subset of “Stress Testing”

• Performed by increasing # of users suddenly

• How application behaves now?

• Measure system performance and server resources

• Similar tools are used as for Load or Stress Testing

Performance Testing – Steps
Tune System (optimization changes)

Analyze Results

Execute Test Revisit Test

Cases/Scripts
Create Test Environment

Create Test Scripts

Choose Suitable Tool

Determine Key Measures (e.g. ordering)

Identify Business Goals