Our great sponsors
-
metering-operator
The Metering Operator is responsible for collecting metrics and other information about what's happening in a Kubernetes cluster, and providing a way to create reports on the collected data.
-
kubeplus
Kubernetes Operator to create multi-instance SaaS from Helm charts using Kubernetes-native APIs
-
InfluxDB
Power Real-Time Data Analytics at Scale. Get real-time insights from all types of time series data with InfluxDB. Ingest, query, and analyze billions of data points in real-time with unbounded cardinality.
Metering Operator. But looks like it's not under active development.
You can check out our project KubePlus (https://github.com/cloud-ark/kubeplus) . It addresses exactly the type of requirements you have. KubePlus collects CPU, memory, storage, network ingress/egress metrics at Helm chart level and exposes them in Prometheus format. In your scenario, the data pipeline and the managed ML models would be the Helm charts. KubePlus provides a way to wrap these behind Kubernetes style APIs (custom resources). The customer facing web app will create instances of these custom resources for data pipeline / managed ML instances. Behind the scene you as the cluster administrator would be able to track CPU, memory, storage, network for each instance of this custom resource (essentially a helm release). Here is an example of Wordpress service delivered using KubePlus https://cloudark.io/helm-chart-as-a-service You can try out the Wordpress example by following these steps: https://cloud-ark.github.io/kubeplus/docs/html/html/getting-started.html The consumption metrics for Wordpress example are seen here: