Agile Operations or How to sleep better at night
5 likes3,227 views
1) Agile values and practices can be applied to operations by focusing on individuals and interactions, comprehensive automation through infrastructure as code, and responding quickly to changes rather than following a predefined plan. 2) Typical system administration involves manual and repetitive installation and configuration tasks, whereas automating infrastructure build through version control and continuous deployment allows infrastructure to be provisioned and configured in a consistent, automated way. 3) Monitoring everything from end-user experience to applications, databases, and operating systems enables detecting problems early. Automated alerts and root cause analysis facilitate fast problem resolution.
Agile Operations or How to sleep better at night
- 1. Agile Operations or How to sleep better at night Raimonds Simanovskis
- 2. @rsim github.com/rsim Raimonds Simanovskis
- 3. The easiest Business Intelligence tool on the Web
- 5. Before Agile... Customer Developer Tester
- 6. Agile Cross-Functional Team Analyze Design Code Test
- 7. Development vs Operations Agile SysAdmins Support
- 8. DevOps Customer Support Developer SysAdmin Tester DBA
- 9. How to apply Agile values and practices to Operations?
- 10. Agile Values Individuals and Processes over interactions and tools Working Comprehensive over production system documentation Customer SLA over collaboration negotiation Responding to Following over change a plan
- 11. Infrastructure as code
- 13. Typical system administration App1 Pkg2 Pkg1 DB OS Installation Development instructions server App1 App1 Pkg2 Pkg3 Pkg2 Pkg3 DB Pkg1 Pkg1 DB OS OS1 OS2 Production Test ser vers server
- 14. Typical system administration App1 Pkg2 Pkg1 DB OS Installation Development instructions server App1 server Pkg2 Pkg3 Test Pkg1 DB OS1 OS2 Production ser vers
- 15. Automate infrastructure build Version control system
- 16. Automate infrastructure build Development sand-box Test server Version control system Production servers
- 17. Local sand-box tools Vagrant + VirtualBox
- 18. Infrastructure provisioning and configuration tools Sprinkle
- 20. From development to production Development Operations Production
- 21. From development to production Development Operations Production
- 22. From development to production Development Operations Production
- 23. Fear of change
- 24. From Big Releases to small deployments Faster feedback Problems localized faster Reduces risk Reduces overhead
- 25. Getting started Continuous integration Automated deployment Real-time alerts Root cause analysis
- 26. Good practices Zero-downtime deployments Feature flags Gradual rollouts A/B split testing
- 28. What to monitor? Everything CPU Memory is fine! Disk Network
- 29. Start with end-user experience HTTP Remote Is it alive? user Time to respond? Our One-time failure or server frequent failures?
- 31. Preventive error log analysis
- 32. What to do when something is wrong? Application 1 Monitoring Application 2 application Database Operating system
- 33. What to do when something is wrong? Application 1 Monitoring Application 2 application Database Operating system Alert
- 34. What to do when something is wrong? Application 1 Monitoring Application 2 application Database Operating system Alert
- 35. What to do when something is wrong? a rt ! Application 1 Re s t Monitoring Application 2 application Database Operating system
- 36. Fault tolerant systems Design for failure Fail fast Collect failure data Restore to known state
- 37. Monitoring tools
- 38. Managing infrastructure with code is fun! Pkg2 Pkg1 DB App1 OS