imay opened a new issue #676: New SQL Optimizer URL: https://github.com/apache/incubator-doris/issues/676 Now, Doris's query optimizer is inherited from Apache Impala. Current optimizer is based on rule and not a cost based optimizer, as a result for some query it can't get optimal execution plan. Besides this, this optimizer is not extensible, so it's difficult to add new operators. I want to implement a new optimizer to replace current one. Now cascades like optimizer has been proven to be extensible and effective in many Database System, such as SQL Server. And there is enough related materials to help to understand it. So I want to implement a optimizer based on cascades. Because query optimizer is a huge project, I can't implement a perfect optimizer at once. I decided to divide this project to several steps. 1. Implementing optimizer framework according to cascades. After this work, the new optimizer can optimize simple query and get its optimal plan. For those queries which it can't optimize we can use old optimizer. 2. Adding more function to replace old optimizer completely. During all of this work, something we should follow. 1. Don't modify any exist class for not including any error posiblity. 2. Be compatible with current plan, make it easy to fallback to old optimizer.
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
