EmmyMiao87 opened a new issue #6483:
URL: https://github.com/apache/incubator-doris/issues/6483


   # Background
   
   Problems with existing optimizers:
   1. The degree of cohesion is not high enough, and there are certain 
limitations on scalability.
   2. Forming from the bottom to the top at one time, instead of dynamic 
planning, the estimation of multiple plans is selected.
   3. The optimization strategy is based on heuristics, and the optimization 
strategy related to data dependence is not handled well.
   
   # New Optimizer
   1. Build the New Optimizer architecture based on the Cascades theory, with 
clear layers and support for dynamic planning. Solve poor scalability and 
support best plan in search space.
   2. CBO, an optimizer based on cost, can filter out the best plan through 
statistical information. Manual gear changes to automatic gear. Solve 
data-dependent optimization.
   
   # Participation
   1. Call for a name for New Optimizer
       Requirements: Can highlight certain characteristics of New Optimizer, 
such as the fastest and most accurate ability to locate the best plan.
   2. We sincerely invite enthusiastic young people who have ideas about New 
Optimizer to work together.
       Requirements: Have a strong passion for optimizer technology. Understand 
Cascades theory, or have other open source product optimizer development 
experience.
       Novice task: Find an operator at random, such as AGG node, determine the 
main data amount and cost (memory, etc.) when executing this operator, and 
further compare the cost model of Greenplum and Postgres System.
   
   # Contact me
   1. Leave your thoughts under this ISSUE
   2. Give to subscribe and send email to [email protected]
   3. Email me: [email protected]


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]



---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to