Database in DevOps
This document discusses challenges with database deployments in DevOps workflows. It notes that 92% of respondents in a 2019 survey saw databases as a bottleneck in DevOps processes. Database updates are often tied to application changes, slowing full deployment times. Most DevOps teams find continuous database deployment difficult. There are also often tensions between DBAs and developers. The document provides some methods to help, such as spreading database knowledge across teams and giving developers more visibility and ownership of production databases and performance. However, fully achieving this integration is seen as risky and difficult. DevOps requires shared ownership of the entire software development and operations lifecycle between all roles.
Database in DevOps
- 1. Database in DevOps Edward Kuo
- 4. 2019 State of Database Deployments in Application Delivery report，92%認為 Database 佈 署是DevOps流程中的瓶頸站
- 8. 在組織上，DBA與開發團隊常常也是呈現對立方， Oracle Tom Kyte (資料庫教父之一)，在 2003 年曾經提出消除開發人員和 DBA 之間圍籬的必要性
- 11. 一些方法
- 12. Forwards and backwards application/schema version compatibility Developers have immediate visibility into queries in production Developers own the full app lifecycle including production query performance. Developers own or participate in database performance incident response Database-specific skill, knowledge, and processes are spread around the team DBAs are shielded from most incidents caused by applications or customers DBAs are focused on proactive, strategic architectural/platform/product initiatives
- 16. 那還能怎樣做
- 19. 不算成功的經驗分享
- 29. Summary
- 34. DevOps is not just automation One of the most unfortunate incarnations of DevOps we see, particularly in large companies, is treating it purely as “Developer Operations,” focused entirely on the care and feeding of CI/CD pipelines. Build engineers bridge the gap between development and production, and so it’s understandable how we ended up here—yet this is not DevOps..