pykorm - Python Kubernetes Object-relational mapping (ORM)

pykorm is a simple library that links your models to their kubernetes counterpart.

Each model and instance on your code is thus directly linked to your kubernetes cluster and modifications are thus reflected both ways.

Examples

Namespaced Custom Resource

Setup

First of all, you need to have Custom Resource Definitions on your cluster.
This README will use the following Namespaced resource. You can apply it on your cluster with kubectl.

apiVersion: apiextensions.k8s.io/v1
kind: CustomResourceDefinition
metadata:
  name: peaches.pykorm.infomaniak.com
spec:
  group: pykorm.infomaniak.com
  names:
    kind: Peach
    listKind: PeachList
    plural: peaches
    singular: peach
  scope: Namespaced
  versions:
  - name: v1
    served: true
    storage: true
    schema:
      openAPIV3Schema:
        type: object
        properties:
          spec:
            type: object
            properties:
              variety:
                type: string
            required:
              - variety


    additionalPrinterColumns:
    - name: Variety
      type: string
      description: The variety of the peach
      jsonPath: .spec.variety

Class definition

In order to link a python class to a kubernetes CustomResourceDefinition, you need to inherit the class from pykorm's NamespacedModel or ClusterModel and annotate it with the kubernetes CRD information like so:

import pykorm

@pykorm.k8s_custom_object('pykorm.infomaniak.com', 'v1', 'peaches')
class Peach(pykorm.NamespacedModel):
    variety: str = pykorm.fields.Spec('variety')

Notice that a class inheriting from pykorm.NamespacedModel already has the name and namespace fields setup.

Create a CR

In order to create a kubernetes custom resource from python, you just have to instantiate the class and save it with Pykorm.save():

import pykorm
pk = pykorm.Pykorm()

cake_peach = Peach(namespace='default', name='cake-peach', variety='Frost')
pk.save(cake_peach)  # We save the resource

as you can see, the model is instantly ensured in kubernetes:

$ kubectl get peach -n default
NAME         VARIETY
cake-peach   Frost

List resources

Pykorm can also list resources from kubernetes

>>> all_peaches = Peach.query.all()
>>> for peach in all_peaches:
>>>  print(peach)
<Peach namespace=default, name=cake-peach, variety=Frost>

# Filter by namespace
>>> Peach.query.filter_by(namespace='default').filter_by(variety='Frost').all()

You can even filter resources by some criterion:

>>> Peach.query.filter_by(name='cake-peach').all()
[<Peach namespace=default, name=cake-peach, variety=Frost>]
>>> Peach.query.filter_by(namespace='kube-system').all()
[]

Delete resources

You can delete a resource with pykorm too:

pk.delete(peach)

$ kubectl get peach
No resources found in default namespace.

More examples

For more examples, don't hesitate to look into the examples/ directory

• CoreAPI Wrapper
• Multi Layer Nested & Mixin Fields
• Label Filter & Field Selector
• Multi Cluster

Is pykorm stable ?

pykorm is still very young and very naive. It's also missing quite a lot of features (relationships, etc.). It was originally created because a lot of boilerplate code was written each time a kubernetes custom object had to be interfaced with python code.

Work on pykorm is actually on the way. Don't hesitate to contribute to the project if you have the energy for it !

Equivalences