Krogon

Krogon is a DSL that encodes repeatable patterns we observed while using K8s. Krogon also has plugins that encode repeatable patterns in services that run on K8s.

Requirements

• python3

Plugins

• Krogon-GoCD
• Krogon-Istio

Installation

Install python3:

To use pyenv to install python, run:

brew install pyenv
pyenv install 3.7.4
eval "$(pyenv init -)" # put in your rc file (e.g. ~/.zshrc)

Setup your virtual environment:

pip install virtualenv
virtualenv .venv  --no-site-packages
. ./.venv/bin/activate

Install Krogon:

pip install -e "git+ssh://git@github.com/enamrik/krogon.git#egg=krogon"

Usage

DSL

Krogon is a simple step runner. There are two steps as of now: a template generator and a template runner.

Template Generator Step

gen_template:

a template generator. Templates are put in a folder called output in the current directory.

e.g.

from krogon import gen_template

gen_template(
    templates = [
    ...
])

Args:

• templates: An array of templates for which to generate yaml files.

The template will be written to <PWD>/output/k8s.yaml.

Template Runner Step

run_in_cluster:

a template runner that executes templates in one or more clusters.

e.g.

import os
from krogon import run_in_cluster, gke_conn

run_in_cluster(
    conn=gke_conn(cluster_regex='prod-us.*', 
                  project_id=os.environ['PROJECT_ID'],
                  service_account_b64=os.environ['SERVICE_ACCOUNT_B64']),
    templates = [
        ...
    ]
)

Args:

• templates: An array of templates for which to generate yaml files.

• gke_conn:cluster_regex: A tag used to filter which cluster(s) to deploy to. In the example, any cluster which name contains the text 'prod-us', e.g. 'prod-us-east' or 'prod-us-west', will be deployed to. Can also be set via environment variable: KG_CLUSTER_REGEX=prod-.*.

• gke_conn:project_id: The project_id of the project where your clusters live. Can also be set via environment variable: KG_PROJECT_ID=prod-project.

• gke_conn:service_account_b64: A base 64 string of the service account key file. Can be produced by $(base64 ~/Documents/prod-access.json). Can also be set via environment variable: KG_SERVICE_ACCOUNT_B64=$(base64 <path-to-file>).

Optional Env Args:

• KG_TEMPLATE: If True, wil generate each cluster's yaml file in the output folder. Default is false.

• KG_DELETE: If True, will delete the resources instead of created.

The conn can also be dynamically discovered:

e.g.

from krogon import run_in_cluster, discover_conn

run_in_cluster(
    conn=discover_conn(),
    templates = [
        ...
    ]
)

If KG_SERVICE_ACCOUNT_B64 is set, gke_conn will be created. If no conn can be determined, the local_kubectl_conn is assumed. local_kubectl_conn used the system kubectl. Krogon will thus deploy to whatever context the kubectl context is pointing to. discover_conn is the default if conn is left out.

When using discover_conn, conn types can only be configured by their environment variable options.

Template types

micro_service:

a template for generating resources that make up a service.

e.g.

from krogon import gen_template
from krogon.k8s import micro_service

gen_template([
    micro_service(
        name='test-app',
        image='gcr.io/prod-project/test-app:1.0.0',
        app_port=3000)
])

Args:

• name: The name of the service. It will be used to name other resources.

• image: The full image url of the container image.

• app_port: The port the container will be listening on. This is different from the service port.

Optionals:

• with_service_port(port: int): The port the service will be made available on. Default is 80.

• with_replicas(min: int, max: int): The min/max replicas allowed to scale to. Default is 1 to 3.

• with_command(command_args: List[str]): The container args. Default is none.

• with_environment_variable(name: str, value: str): The environment variables set for the container process.

• with_environment_secret(secret_name: str, data: dict): Adds a reference to a secret. secret_name is the name of a secret and data is a map where key is the environment variable name and value is the key of the secret part.

secret:

a template to generate a secret.

Args:

• name: The name of the secret.

• data: A map of the key and value of key parts.

• already_b64: True if the value of key parts don't need to be base64 encoded. Default is false.

cron_job:

a template for creating cron jobs.

• name: The name of the job.

• image: The full image url of the container image.

Optionals:

• with_command(command_args: List[str]): The container args. Default is none.

• with_environment_variable(name: str, value: str): The environment variables set for the container process.

• with_environment_secret(secret_name: str, data: dict): Adds a reference to a secret. secret_name is the name of a secret and data is a map where key

• with_schedule: The cron pattern that determines how often the job runs. The default is * * * * * (disabled).