[Feature] Multi-Cluster Information Collector

[Somefive]

What we need

With the advancement of ClusterGateway, we should support providing more cluster information, such as the health status, available cpu cores, whether it is a GPU cluster, the stability of connections with KubeVela control plane, and so on. Generally, the cluster information could be highly extensible.

Benefits of having it

With these provided cluster information, more fancy upper mechanisms could be supported, such as dynamic workload scheduling, monitoring metrics exposure, or system failure diagnoses.

How to do

Push mode

A lightweight operator (or synchronizer) is needed to do the job: watch every managed cluster and retrieve metrics (or information) from time to time.

Pull mode

An alternative way is to install agents in every managed cluster. These agents are responsible for retrieving metrics (or information) and expose them (or report them). This solution is heavier than the push mode but more scalable and efficient.

[JooKS-me]

@Somefive @wonderflow Hi, I am a junior in NEU, I'm interested in this issue and want to get mentored in LXF Mentorship, so I have submitted an application in the LXF system. I have made observability related contributions to some open source projects, such as Apache ShardingSphere, ShenYu. After a few days of study, I have a lot of understanding and thinking about KubeVela, and I hope to contribute to this community~