Application Monitoring

The monitoring attribute in the AppConfiguration instance is used to describe the specification for the collection of monitoring requirements for the application.

As of version 0.9.0, Kusion supports integration with Prometheus by managing scraping behaviors in the configuration file.

info

The monitoring attribute requires the target cluster to have installed Prometheus correctly, either as a Kubernetes operator or a server/agent.

More about how to set up Prometheus can be found in the Prometheus User Guide for Kusion

Import

In the examples below, we are using schemas defined in the catalog package. For more details on KCL package import, please refer to the Configuration File Overview.

The import statements needed for the following walkthrough:

import catalog.models.schema.v1 as ac
import catalog.models.schema.v1.workload as wl
import catalog.models.schema.v1.monitoring as m

Project-level configurations

In addition to the KCL configuration file, there are also project-level configurations that can be set in the project.yaml in hte project root directory.

By separating configurations that the developers are interested in and those that platform owners are interested in, we can reduce the cognitive complexity of the application configuration and achieve separation of concern.

In the context of monitoring, there are two flags you can set in project.yaml that will alter the behavior of Kusion.

info

If you have initialized the projects with kusion init, the project.yaml should be automatically created for you.

Operator mode

The operatorMode flag indicates to Kusion whether the Prometheus instance installed in the cluster runs as a Kubernetes operator or not. This determines the different kinds of resources Kusion manages.

To see more about different ways to run Prometheus in the Kubernetes cluster, please refer to the design documentation.

Most cloud vendors provide an out-of-the-box monitoring solutions for workloads running in a managed-Kubernetes cluster (EKS, AKS, etc), such as AWS CloudWatch, Azure Monitor, etc. These solutions mostly involve installing an agent (CloudWatch Agent, OMS Agent, etc) in the cluster and collecting the metrics to a centralized monitoring server. In those cases, you don't need to set operatorMode to True. It only needs to be set to True when you have an installation of the Prometheus operator running inside the Kubernetes cluster.

info

For differences between Prometheus operator, kube-prometheus and the community kube-prometheus-stack helm chart, the details are documented here.

Monitor types

The monitorType flag indicates the kind of monitor Kusion will create. It only applies when operatorMode is set to True. As of version 0.9.0, Kusion provides options to scrape metrics from either the application pods or its corresponding Kubernetes services. This determines the different kinds of resources Kusion manages when Prometheus runs as an operator in the target cluster.

A sample project.yaml with Prometheus settings:

# The project basic info
name: multi-stack-project
generator:
  type: AppConfiguration
prometheus:
  operatorMode: True
  monitorType: Service

To instruct Prometheus to scrape from pod targets instead:

# The project basic info
name: multi-stack-project
generator:
  type: AppConfiguration
prometheus:
  operatorMode: True
  monitorType: Pod

If the prometheus section is missing from the project.yaml, Kusion defaults operatorMode to false.

Managing Scraping Configuration

To create scrape configuration for the application:

myapp: ac.AppConfiguration {
    workload: wl.Service {
        # ...
    }
    monitoring: m.Prometheus{
        interval:       "30s"
        timeout:        "15s"
        path:           "/metrics"
        port:           "web"
        scheme:         "http"
    }
}

The example above will instruct the Prometheus job to scrape metrics from the /metrics endpoint of the application every 30 seconds.

To instruct Prometheus to scrape from actuator/metrics on port 9099 instead:

myapp: ac.AppConfiguration {
    workload: wl.Service {
        # ...
    }
    monitoring: m.Prometheus{
        interval:       "10s"
        timeout:        "5s"
        path:           "/actuator/metrics"
        port:           "9099"
        scheme:         "http"
    }
}

More details about how the Prometheus integration works can be found in the design documentation.