[Feature] VelaQL for Observability

[Somefive]

What we need

VelaQL in KubeVela is powerful and currently is widely used in the implementation of api-server interfaces for VelaUX. It is easy for user to write a single line VelaQL sentence to query the pods behind an application or even collecting logs inside. Now user can only leverage this capability through KubeVela api-server. If we can integrate it into observability system, it would benefit the monitoring system and help diagnose problems more conveniently.

To be precise, this feature want to provide a way for user to use VelaQL to query results and view them on Grafana dashboard.

Benefits of having it

Users can observe more customized information inside KubeVela system through commonly used observability tools like Grafana or Prometheus. As the VelaQL is highly configuration, it is also convenient for user to extend it in the Kubernetes eco-system.

How to do

Extend with Prometheus Metrics

Develop a metrics reporter which can evaluate VelaQL queries routinely and expose the results through Prometheus Metrics. Then Prometheus will discover it and then Grafana display will use the Prometheus metrics.

Direct Data Source Integration

An alternative way is to extend a data source in Grafana which allows user to write VelaQL queries directly inside grafana and retrieve results.