Utilizing Containers for Performance and Scale Engineering
0 likes193 views
Utilizing containers allows for easier performance and scale testing of complex software architectures. Containers make it simple to replicate real-world environments and monitor components at scale. Red Hat uses containers to test Satellite 6, which provides provisioning, configuration management, and software updates across platforms. Their containerized performance tests allow thousands of Satellite instances to run on a single machine for testing and optimization.
Utilizing Containers for Performance and Scale Engineering
- 1. Utilizing Containers For Performance and Scale Engineering Saurabh Badhwar Associate Software Engineer, Performance and Scale Engineering Red Hat
- 2. Software Architectures are Complex, their performance and scale analysis... Even more
- 3. But Why?
- 4. Number of Interacting Components Reverse Proxy Python WSGI App Ruby Application Database Database Message Bus Foreman Project with Katello Plugin
- 5. Benchmarking in Isolation Doesn't Provide Real Numbers Ruby Application Integrated Product 10000+ connections handled with breeze ~1000 connections handled
- 6. Hard To Replicate Real World Use Cases Content Download Authentication Configuration Management Messaging Security And ?
- 7. Solving our problems using containers
- 8. Software Deployment Made Easy Operating System Software Environment
- 9. Easy To Replicate User Environments Operating System Software Environment (Server) Client System Client System Client System
- 10. Easy To Monitor Components Operating System Software Environment (Server) Client System Client System Client System Monitoring Server Monitoring Backend Monitoring Frontend Carbon+Graphite Grafana
- 11. Any Real Life Examples?
- 12. Red Hat Satellite 6 A case study
- 13. Red Hat Satellite 6 What it does? Provisioning Configuration Management Software Management Subscription Management All This Across a variety of Platforms Bare Metal Systems Virtual Machines Containers
- 14. Red Hat Satellite 6 Architecture
- 15. Red Hat Satellite 6 Why we care about Performance and Scale? To understand how our software affects user experience To find the extent till where we can go To provide optimizations and alternate use cases To rule out bugs at scale
- 16. Red Hat Satellite 6 How we are using Containers in Performance and Scale? Satellite 6 Server Monitoring Server Atomic Host RHEL 7.3 Container 1 RHEL 7.3 Container 2 RHEL 7.3 Container 3 RHEL 7.3 Container 4 RHEL 7.3 Container 1000 collectd
- 17. Time for a demoLet's see how it all looks like
- 18. Benefits of Using Containers For Performance and Scale Engineering
- 19. Benefits of Using Containers Number of instances on same machine
- 20. Benefits of Using Containers Resource usage overhead
- 21. Benefits of Using Containers Time to recreate environment
- 22. Benefits of Using Containers Time to initialize a single instance
- 23. How our Red Hat Satellite Scale Environment Looks like With Containers 1 x Satellite Server 10 x Satellite Capsule Server 32 x Atomic Hosts = ~32k RHEL 7.3 Containers 1 x Monitoring Server
- 24. Thank You :D
- 25. Q & A Reach Out To Me At: E-Mail: sbadhwar@redhat.com Twitter: @MYNET55 LinkedIn: linkedin.com/in/sbadhwar Credits Pradeep Surisetty, Red Hat Jan Hutar, Red Hat