Kubernetes Schedule Scaler

Kubernetes Schedule Scaler allows you to change the number of running replicas of a Deployment at specific times. It can be used is to turn on/off applications that don't need to be always available and reduce cluster resource utilization, or to adjust the number of replicas when it's known in advance how the traffic volume is distributed across different time periods.

Installation

$ kubectl apply -f deploy/deployment.yaml

Usage

Just add the annotation to your Deployment:

  annotations:
    hoopla/scaling.schedule: '[{"schedule": "10 18 * * *", "replicas": "3"}]'

The following fields are available:

• schedule - cron expression for the schedule
• replicas - the number of replicas to scale to

The value of replicas can also be a pointer to an annotation value on the same deployment. The final value will be copied from the annotation. See usage example for annotation pointers below.

Deployment Example

kind: Deployment
metadata:
  name: nginx-deployment
  labels:
    application: nginx-deployment
  annotations:
    some-annotation-value: "3"
    other-annotation-value: "0"
    hoopla/scaling.schedule: |
      [
        {"schedule": "0 7 * * MON-FRI", "replicas": "1"},
        {"schedule": "0 19 * * MON-FRI", "replicas": "other-annotation-value"},
        {"schedule": "0 12 * * MON-FRI", "replicas": "5"},
        {"schedule": "0 16 * * MON-FRI", "replicas": "some-annotation-value"}
      ]

Temporarily disabling scheduled scaling

You can temporarily disable scheduled scaling by adding an annotation hoopla/scaling.disabled with the value "true".

kind: Deployment
metadata:
  name: nginx-deployment
  labels:
    application: nginx-deployment
  annotations:
    hoopla/scaling.disabled: "true"
    hoopla/scaling.schedule: |
      [
        {"schedule": "0 7 * * MON-FRI", "replicas": "1"},
        {"schedule": "0 19 * * MON-FRI", "replicas": "0"}
      ]

Predefined/shared schedules

Configure predefined schedules by passing the environment variable PREDEFINED_SCHEDULES to schedule_scaling/main.py with a json object. E.g.:

{
  "predefined-schedule-a": [
    {"schedule": "0 7 * * MON-FRI", "replicas": "1"},
    {"schedule": "0 19 * * MON-FRI", "replicas": "0"}
  ],
  "predefined-schedule-b": [
    {"schedule": "* 0/2 * * *", "replicas": "1"},
    {"schedule": "* 1/2 * * *", "replicas": "0"}
  ]
}

Annotation pointers are still fetched from the deployment, so it's possible to have different replicas values for different deployments even when using predefined schedules.

In the deployment annotation, set the annotation hoopla/scaling.schedule.predefined instead of hoopla/scaling.schedule.predefined:

kind: Deployment
metadata:
  name: nginx-deployment
  labels:
    application: nginx-deployment
  annotations:
    hoopla/scaling.schedule.predefined: predefined-schedule-a

If both annotations are present, the predefined schedule takes precedence.

Logging

You can change the global log level using the LOG_LEVEL environment variable (e.g. LOG_LEVEL=DEBUG) and the module log level using the SCHEDULED_SCALING_LOG_LEVEL environment variable.

Dry run

Set the DRY_RUN environment variable to true to make the kubernetes commands dry run only.