Optimize resource recommendation controller
pendoragon opened this issue · comments
Why is this needed?
Katalyst has recently added a resource recommendation controller which can work with VPA to help user allocate proper resource request/limit for pods. However, there're still gaps for delivering resource recommendation in a release:
- Now we build the Controller using the Kubernetes Controller-Runtime Project. while Katalyst has its own controller framework(e.g. vpa controller). We expect to use this framework to refactor the controller section
- It lacks deployment artifacts
- It lacks relevant documentations
What abilities do you need?
- Familiar with golang programming, have the ability to independently analyze and solve problems, and proficient in using search engines.
- Familiar with or willingness to learn the k8s controller-manager mechanism, it would be even better if you are familiar with or have used The Kubernetes controller-runtime Project.
- Familiar with or willingness to learn the usage and code of the Katalyst controller framework(e.g. vpa controller).
- Have some experience in developing and using helm chart
- Have ability to write documents
- Can reasonably arrange time to participate in the design and development of open source projects, and maintain a passion for learning.
What would you like to be added?
Code
- Refactor resource recommendation controller code so that it aligns with other katalyst controllers
- Add helm chart for resource recommendation
Documents
- Add a quick start to gokatalyst.io