This task is to explore the Kuberenetes operator model. There are several operators being developed currently, each with their own characteristics. The goal is to look at operators that are designed to operate software that creates cluster such as Crunchy Data's Postgres and CoreOS' etcd operator.
[x] Explore codebase and api used
[x] Understand the operator pattern
[x] Find out which Kubernetes primitives are used for controllers
[x] Figure out level of effort required to build Operators
[x] Investigate deprecation of TPR (ThirdPartyRes) and its replacement with CRD (custom resource def) in 1.7 - (6/29)
[ ] Explore new server extension models in Kubernetes (api server aggregation, custom resources, etc) (7/7)
[ ] Sketch out minimal requirements for operator to manage clustered storage software (7/8)
[ ] Create a Proof-of-Concept operator to test ideas, understand implications, and date estimates (8/11)
Outcome
High level of comfort writing code the Kubernetes go-client
Design for creating an operator to managed clustered storage software
Develop insights for simplifying steps for creating distributed applications on Kubernetes
This task is to explore the Kuberenetes operator model. There are several operators being developed currently, each with their own characteristics. The goal is to look at operators that are designed to operate software that creates cluster such as Crunchy Data's Postgres and CoreOS' etcd operator.
Outcome