Some of you have seen my presentation about hidden parts of Oracle execution plans and how to access the plan in the memory directly and parse it. I presented it at OakTable World 2017 and it will be also presented in Wellington and Acukland this November. You can download the presentation at http://vitspinka.com/files/ReadingPlanFromSGA-OTWatOOW-2017.pdf.
I realize that many people did not have the chance to attend... and that the slides need quite a lot of explaining, it's hard to understand this internals without more explanation.
Thus you can look forward to seeing a handful of blog posts, inspired by this presentation, and explain some of the aspects of this whole topic.
We will start with some basic tools; accessing SGA and PGA, which you may find useful for many other tasks, too. Then we will look at some details of the execution plan. This is not - and neither the presentation is - an exhaustive guide to the execution plan internals. It would be a multi-year project to test all the possible cases and to implement a reliable parser. Instead, I want to focus on the principles - how to approach the tasks, and the major principles I've learned about the execution plan.
I realize that many people did not have the chance to attend... and that the slides need quite a lot of explaining, it's hard to understand this internals without more explanation.
Thus you can look forward to seeing a handful of blog posts, inspired by this presentation, and explain some of the aspects of this whole topic.
We will start with some basic tools; accessing SGA and PGA, which you may find useful for many other tasks, too. Then we will look at some details of the execution plan. This is not - and neither the presentation is - an exhaustive guide to the execution plan internals. It would be a multi-year project to test all the possible cases and to implement a reliable parser. Instead, I want to focus on the principles - how to approach the tasks, and the major principles I've learned about the execution plan.
Comments