Skip to content

Latest commit

 

History

History
138 lines (95 loc) · 4.94 KB

README.md

File metadata and controls

138 lines (95 loc) · 4.94 KB
Raw

icon

Kronic

Build / Test License

The simple Kubernetes CronJob admin UI.

Kronic is in early alpha. It may eat your cronjobs, pods, or even your job. Avoid exposing Kronic to untrusted parties or networks or using Kronic near anything even vaguely important.

Screenshots

See CronJobs across namespaces: Homepage

View, suspend, trigger, clone, or delete CrobJobs at a glance: Cronjobs in a Namespace

Drill down into details to see the status of jobs and pods: Cronjob Detail view

Get your hands dirty with the raw YAML to edit a CronJob: Cronjob Edit view

Purpose

CronJobs are a powerful tool, but I have found that developers and stakeholders often need an easy way to inspect the status of jobs, trigger them ad-hoc, or create a new one-off job based on existing CronJob definitions.

Kronic aims to be a simple admin UI / dashboard / manager to view, suspend, trigger, edit, and delete CronJobs in a Kubernetes cluster.

Configuration

Kronic can be limited to a list of namespaces. Specify as a comma separated list in the KRONIC_ALLOW_NAMESPACES environment variable. The helm chart exposes this option.

Kronic also supports a namespaced installation. The KRONIC_NAMESPACE_ONLY environment variable will limit Kronic to interacting only with CronJobs, Jobs and Pods in its own namespace. Enabling this setting in the helm chart values (env.KRONIC_NAMESPACE_ONLY="true") will prevent creation of ClusterRole and ClusterRolebinding, creating only a namespaced Role and RoleBinding.

Authentication

Kronic supports HTTP Basic authentication to the backend. It is enabled by default when installed via the helm chart. If no password is specified, the default username is kronic and the password is generated randomly. A username and password can be set via helm values under auth.adminUsername and auth.adminPassword, or you may create a Kubernetes secret for the deployment to reference.

To retrieve the randomly generated admin password:

kubectl --namespace <namespace> get secret <release-name> -ojsonpath="{.data.password}" | base64 -d

To create an admin password secret for use with Kronic:

kubectl --namespace <namespace> create secret generic custom-password --from-literal=password=<password>

## Tell the helm chart to use this secret:
helm --namespace <namespace> upgrade kronic kronic/kronic --set auth.existingSecretName=custom-password

Deploying to K8S

A helm chart is available at ./chart/kronic. By default the Kronic helm chart will provide only a ClusterIP service. See the values.yaml for settings, most notably the ingress section.

Warning Avoid exposing Kronic publicly! The default configuration allows for basic authentication, but provides only minimal protection.

To install Kronic as kronic in its own namespace:

helm repo add kronic https://mshade.github.io/kronic/
helm repo update

# Optionally fetch, then customize values file
helm show values kronic/kronic > myvalues.yaml

helm install -n kronic --create-namespace kronic kronic/kronic

# See the NOTES output for accessing Kronic and retrieving the initial admin password

If no ingress is configured (see warning above!), expose Kronic via kubectl port-forward and access localhost:8000 in your browser:

kubectl -n kronic port-forward deployment/kronic 8000:8000

Running Locally

Kronic can use a KUBECONFIG file to run locally against a cluster. To do so:

docker run -i --name kronic \
    -v $HOME/.kube:/home/kronic/.kube \
    -p 8000:8000 \
    ghcr.io/mshade/kronic

Note You may need to ensure permissions on the kubeconfig file are readable to the kronic user (uid 3000). You may also mount a specific kubeconfig file into place, ie: -v $HOME/.kube/kronic.yaml:/home/kronic/.kube/config

Design

Kronic is a small Flask app built with:

Todo

  • CI/CD pipeline and versioning
  • Helm chart
  • Allow/Deny lists for namespaces
  • Support a namespaced install (no cluster-wide view)
  • Built-in auth option
  • Display elements and handling for spec.timezone
  • NetworkPolicy in helm chart
  • Timeline / Cron schedule interpreter or display
  • YAML/Spec Validation on Edit page
  • Async refreshing of job/pods
  • Error handling for js apiClient
  • Better logging from Flask app and Kron module
  • Improve test coverage
  • Integration tests against ephemeral k3s cluster
  • Improve localdev stack with automated k3d cluster provisioning