Performance Testing : Cloud Deployments
•
1 like•124 views
Delivered talk on Performance Testing : Cloud Deployments by Shreyas Chaudhari and Manish Hemnani at ThoughtWorks, Pune on 16th March, 2019 in VodQA Pune 2019.
Report
Share
Performance Testing : Cloud Deployments
- 1. Performance Testing Cloud Deployments Shreyas & Manish vodQA, Pune 16th March 2019
- 2. Shreyas Chaudhari Quality Analyst at ThoughtWorks, Inc About Me
- 3. Manish Hemnani About Me Quality Analyst at ThoughtWorks, Inc
- 4. 4 Agenda ● Why Performance Testing ● Challenges ● Testing - Cloud Vs On-Premise ● Why AWS? ● Elastic Load Balancing [ELB] ● Auto Scaling ● Case Study - Architecture, Test Strategy ● AWS Console ● CloudWatch ● New Relic
- 6. 6 Solution to Performance Issues
- 7. 7 Patterns
- 8. 8 Fluctuations - Different times of day
- 9. 9 Fluctuations - Different days of year
- 13. 13 Pros & Cons Source : https://www.cms-connected.com/getattachment/News-Archive/November-2017/Cloud-Digital-Asset-Man agement/Cloud-Vs-On-Premises.jpg
- 15. 15 Why AWS?
- 16. 16 Public Cloud Adoption Source : https://www.datamation.com/imagesvr_ce/7252/aws-azure-google.png
- 18. 18 Elastic Load Balancing [ELB]
- 19. 19 Cross Zone Load Balancing - Enabled https://docs.aws.amazon.com/elasticloadbalancing/latest/userguide/images/cross_zone_load_balancing_enabled.png
- 20. 20 Cross Zone Load Balancing - Disabled https://docs.aws.amazon.com/elasticloadbalancing/latest/userguide/images/cross_zone_load_balancing_disabled.png
- 21. 21 Auto Scaling
- 22. 22 What is Auto Scaling? https://docs.aws.amazon.com/autoscaling/ec2/userguide/images/as-basic-diagram.png
- 23. 23 Auto Scaling Options Maintain current instance levels all the time Manual scaling Scaling based on a schedule Scale based on demand
- 24. 24 Availability
- 26. 26 Performance Testing - Case Study
- 28. 28 Test Environment Configuration Component Instance Type Number of Instances Scooper t2.medium 1 EventStore t2.large 3 Subscriber(s) t2.medium 1 Elastic Search m4.4xlarge.elasticsearch Master : 3 Data : 4 MongoDB t3.large 3 Service(s) t3.xlarge 3 Web t3.medium 3 Instance Types : https://aws.amazon.com/ec2/instance-types/
- 29. 29 Testing Types Load Stress Spike Endurance Page Load
- 30. 30 Goal 99th percentile Page load time of the application for the end user from different geographies should be 2 seconds or less. ‘N’ Concurrent users should be able to use the application
- 31. 31 Test Strategy - Prerequisites Perform Writes on Elastic Search and MongoDB using Data Migration Scripts Perform Read on Elastic Search and MongoDB using Apis
- 32. 32 Tools
- 34. 34 AWS Console - Parameters
- 38. 38 MongoDB & Public Apis
- 40. 40 CloudWatch Source : https://eu-west-2.console.aws.amazon.com/cloudwatch/home?region=eu-west-2
- 41. 41 New Relic Source : https://rpm.newrelic.com/accounts/1570350/applications/192890835
- 42. 42 New Relic Source : https://rpm.newrelic.com/accounts/1570350/applications/192890835/transactions
- 43. 43 New Relic Source : https://insights.newrelic.com/accounts/1570350/dashboards/845142
- 44. THANK YOU For questions or suggestions mailto: shreyasc@thoughtworks.com hmanish@thoughtworks.com