Strategies to Edit Production Data
summary
Five concrete strategies for making production writes safer: raw SQL with a Google spreadsheet audit trail, local Python scripts with dry‑run flags, remote execution on existing servers, Jenkins as a task runner with enforced code review and persistent logs, and a dedicated Script Runner service with centralized credentials and environment segregation. Examples include converting queries into reusable Python scripts, hitting a 50 million‑row update pain point, CPU contention on Jenkins, and how Script Runner fixes argument validation, auditing, and long‑running jobs.
AI-written summary · about
chapters (11)
- 0:00Introduction to production data editing
- 1:14Why editing production data is risky
- 5:29Five strategies for safer edits
- 6:32Manual review spreadsheet process
- 11:30Running scripts locally
- 15:29Running scripts on existing servers
- 20:10Using a task runner
- 24:22Building a script runner service
- 27:12Choosing the right strategy
- 28:15Key lessons and takeaways
- 29:19Q&A and closing remarks
related talks
26mStrategies to Edit Production Data
27mStrategies to Edit Production Data
29mStrategies to Edit Production Data
29mStrategies To Edit Production Data
21mPractical Deployment Strategies for Production Python Apps
01h 01mKeep your nose out of it. Denying yourself access to production
30mReliable Data Processing with Minimal Toil
25mOptimize your Data Pipeline without Rewriting it