Performance Engineering – Performance Engineer – Activities

Common problems in Production environment

• When CPU reaches its 100%, we will get RequestTimedOut Errors, by which the response time increases much
• High memory utilization – Elastic Search / Redis caching or API deployed in EC2 instance has generated much logs thereby consuming more memory
• Other issues could be
  • In-Efficient source code that could result in
    • Memory Leaks
    • Dead Lock issues
    • Not using proper data structures that could lead to more CPU an memory usage, etc
  • DB issues
    •  Issues with indexes, etc
  • Not setting up of infrastructure to suite to future growth rate(users, transactions, etc) of data
  • Peak application access on specific days – Thanks Giving, etc
  • Improper load distribution
  • Issues with DNS, Firewall, Network connectivity
  • Issues with a dependent 3rd party application
  • Issues with shared resources
  • Issues with threads
  • Issues with Garbage Collection Memory issues
    • Memory Leak
    • Not leveraging caching mechanism
    • Not allocating proper memory
  • Domino Effect
    • An issue existing at one place creates issues in other places, and we have to properly backtrack the issue for RCA
  • Other miscellaneous issues
    • Network issues
    • Application doesn’t respond as usual, due to scheduled job
  • It will be better if the performance engineer focuses on upskilling on the below aspects
    • Learning programming languages
    • JVM internals
    • Garbage Collection internals
    • Memory Leak issues
    • Source Code tuning
    • Performance Tuning
    • Performance Testing Tools – Load Runner, Neo Load, jMeter, Gatling, etc
    • Understanding the workload – Little’s Law, etc
    • Analyzing the application’s / platform’s logs
    • Identify critical business flows
    • Test data management
    • Basics of script development
    • Understand the performance metrics, parameters, bottlenecks of the Performance Test Report
    • Clear understanding of their application’s architecture
    • Database tuning
    • Analyze the metrics, information from Application’s monitoring tools – Dynatrace, App Dynamics, New Relic, Coralogix, Data Dog, CA APM, Splunk, etc
    • Good grasp on operating system internals
    • Understanding on Load Balancer internals
    • Awareness towards good understanding on Mobile, Cloud, Containerization, Microservices concepts
    • Understanding on browser side metrics – Single user metrics
    • Code Profiling Tools

-By  Saravanan & Sai Krishna