In one week's time I'll be presenting common_schema: DBA's framework for MySQL at Percona Live, London.
This talk introduces the rich toolset known as common_schema. It is free, open source, extremely useful in solving DBA & developer tasks, and is the next best thing ever invented next to SQL pie charts.
- Views, with which you can detect and terminate idle transactions, blocking transactions; roll your range partitions; find duplicate keys; block/unblock accounts; get fine grained privileges per account; find AUTO_ICNREMENT free space; ...
- Routines: do meta executions such as eval(); get query checksums; duplicating accounts; killing suspicious connections; security auditing; parsing JSON data; ...
- QueryScript: if you're not using it, you're missing on a powerful scripting language tightly integrated with SQL/MySQL. We'll see the basic constructs, variables, loops; the more sophisticated MySQL/locks/overhead/danger aware constructs such as foreach & split; throttling, exceptions, it's all in there. I'll present real scripts that saved the day and challenge you to implement them in another scripting language.
- Briefly introducing rdebug: stored routine debugger and debugging API
- Roadmap (some cool things coming along)
What this talk isn't
A tedious "read the manual aloud". Nor is it a comprehensive listing of all functionality. These would be the surest way of sending you to sleep.
What this talk is
A view into the concepts behind common_schema; the reasons this project has justification to exist; the various problems it solves; and yes, interesting examples. Every single common_schema feature was developed out of real world need.
I am confident you'll find common_schema to have something you need that will improve your work as a developer or a DBA.
The talk is similar to the one I gave at Santa Clara this April. It is updated with new content following recent developments.