Neev Load Testing Services
0 likes729 views
This presentation includes: - Why performance matters for digital businesses? - Use Cases for performance / load testing - Load Test Design Considerations - Tools and Technologies - Methodology and Approach - Activities and Deliverables - Load Testing Success Stories
Ad
More Related Content
What's hot (20)
Similar to Neev Load Testing Services (20)
Ad
More from Neev Technologies (20)
Ad
Recently uploaded (20)
Neev Load Testing Services
- 1. Neev Load Testing Services
- 2. Magento Hybris Commerce SaaS Applications Adobe Marketing Cloud Custom Development Key Company Highlights 300+ team with experience in managing offshore, distributed development. Neev Technologies established in Jan ’05 VC Funding in 2009 By Basil Partners Part of Publicis Groupe Hybris and Adobe CQ centers of Excellence Offices at Bangalore, Gurgaon, Pune, Mumbai Member of NASSCOM About Neev Mobile Cloud iPhone Android PhoneGap Windows Phone HTML5 Apps Web AWS Rackspace Joyent Heroku Google Cloud Platform Digital Marketing, CRM, Analytics (Omni-Channel) User Interface Design and User Experience Design Performance Consulting Practices Quality Assurance & Testing Outsourced Product Development Click here to know more about us
- 3. Document Objectives • About Neev • Why performance matters for digital businesses? • Use Cases for performance / load testing • Load Test Design Considerations • Tools and Technologies • Methodology and Approach • Activities and Deliverables • Load Testing Success Stories
- 4. Why performance matters for digital business? Source: http://www.appdynamics.com/blog/devops/load-testing-tools-explained-the- server-side/ Fast Application Performance Positive / Hassle- Free User Experience Positive Affect towards the Brand Improved Brand Equity A few statistics about the business impact of performance at major internet companies: • Amazon and Walmart increased revenue 1% for every 100ms of improvement • Microsoft found that Bing searches that were 2 seconds slower resulted in a 4.3% drop in revenue per user • When Mozilla shaved 2.2 seconds off their landing page, Firefox downloads increased 15.4% (60 million more downloads) • Making Barack Obama’s website 60% faster increased donation conversions by 14% (30 million more dollars) • Shopzilla sped up average page load time from 6 seconds to 1.2 seconds, and increased revenue by 12% and page views by 25%
- 5. Use cases for performance/load testing • Pre-go-live validation of application’s ability to meet performance SLAs • Benchmark maximum operating capacity (Throughput, concurrent connections, average response time, etc.,) of the system under both steady-state and peak load conditions. • Identify performance bottlenecks on application, database and infrastructure that could potentially cause performance degradations. • Sizing & planning the production infrastructure needed for the system for optimal performance. • Measuring Scaling Limits and Potential Break Points of the system • Continuous Improvement and system maintenance.
- 6. Load Test Design Considerations • Type of Application to be load tested – The load test design may vary depending on the type of application. For example, an application can be any of the following types: internet –based consumer applications, Transactional / workflow based systems, CMS systems, Batch Processing Applications etc., Each of these types of application will require a different approach and thought process when designing the load test. • Typical and Peak Load Expectations: Understanding the expected load traffic under both typical as well as peak load conditions is a key design consideration for load tests. This enables the team to select the right load generator tools for the given application under test. • Baselined Performance Metrics : It is important to lay down clearly what the key performance metrics to be tracked, monitored and base-lined are. This enables the team to select the right monitoring tools for the right data.
- 7. Load Test Design Considerations • Technical and Deployment Architectures: The production architecture of the system under test is a key design consideration. For example various architectural elements such as server clustering, 2 / 3 tier architecture, Use of CDNs, Database replication / sharing set up, etc., will introduce different considerations when designing the load test. • Simulating Geographically distributed user load : Real users may be geographically distributed across different continents all over the world. It is important for the load test to mimic the real world user traffic as closely as possible by generating a geographically distributed load.
- 8. Tools & Platforms 1. Apache Jmeter 2. Amazon Web Services 3. Apache Benchmark (ab) 4. 3rd Party Services (Saucelabs, Blazemeter) 5. APM Solutions (New Relic, AppDynamics) 6. Custom scripts for capturing custom metrics Tools and Technologies
- 9. Methodology and Approach The above time estimates are based on a typical experience. This may vary depending on the exact nature of application involved.
- 10. Deliverables 1. Load Test Plan • Load test objectives, metrics to be captured and test procedures to be executed • Test Infrastructure Plan 2. Load Test Scripts • Recorded and validated load test scripts 3. Load Test Report and Recommendations • Performance benchmark metrics of the application under test • Recommended Infrastructure size for production • Projected Cost of infrastructure and Cost-to-Performance ratio.
- 11. Load Testing Success Stories – KEH About the Client: Founded in the year 1979 in Atlanta, USA, KEH is the world’s largest dealer of used photographic (digital and film) products. KEH lets anyone in any corner of the world sell or trade equipment with a global customer base. Approach: The staging environment served as the test bed for all performance tests conducted in this project. JMeter was used to generate the load required. 1 Amazon EC2 small instance in US Zone acted as the JMeter controller. 5 Amazon EC2 small instances in US Zone acted as JMeter agents. Both the Application Server and the Database server were configured with monitoring systems (Monyog, Manage Engine - App Manager) that captured CPU, Memory, Disk usage, Number of connections etc. Insights: • Even at high load of 37500 users, the application home page responded at an average time of 20ms. • Based on the average response time of 20 ms, we predict that the home page can sustain a load of up to 50,000 users ramping up in 1000 seconds. i.e, up to 50 concurrent users. • There is always a clean throughput. Meaning requests do not queue up at the server end. • The number of open DB connections remain untouched. Meaning, the database Is not impacted by the home page hits. All home page requests are served from the server cache without accessing the database. Recommendations Provided: • Presently an order is at least 10 clicks away for all users. We can look at optimizing the order submit flow. • Review the IIS connections setting and the Database connections setting to tune it according to expected user traffic. • Consider setting Up sufficient monitoring during application failure / ISS server failure.
- 12. The Neev Edge • End-to-end consultative approach for software solutions through needs assessment, process consulting and strategic advice. • Internal QMS are ISO 9001-2008 certified and CMM level 3 compliant. • Continuous process and service level improvements through deployment of best-of-breed processes and technologies. • International Standards and best practices on Project Management including PMI, ISO and Prince-2. • Proven EDC Model of delivery to provide predictable results. • Scrum based Agile development methodology.
- 13. A Few Clients
- 14. Partnerships
- 15. Neev Information Technologies Pvt. Ltd. [email protected] India - Bangalore The Estate, # 121,6th Floor, Dickenson Road Bangalore-560042 Phone :+91 80 25594416 India - Pune Office No. 4 & 5, 2nd floor, L-Square, Plot No. 8, Sanghvi Nagar, Aundh, Pune - 411007. Phone :+91 20 64103338 For more info on our offerings, visit www.neevtech.com