Environment Variables

Environment variables are a powerful way to manage application runtime configurations. Containers running on Cycle can have runtime environment variables injected from a variety of sources. The platform also injects many variables about the container and the context it is running in on Cycle. Environment variables are shared amongst all instances of the container.

User-Defined Environment Variables

Custom user-defined environment variables can be specified several different ways

Image Variables

A container image can define environment variables and their values. Image variables will ALWAYS be injected into the container. They can be overridden on the container runtime configuration, but cannot be removed.

Defining environment variables on an image looks like this:

ENV MY_NAME="John Doe"
ENV MY_DOG=Rex\ The\ Dog
ENV MY_CAT=fluffy

Environment variables can also be set in a Dockerfile without a value, to use as a placeholder when deploying on Cycle.

ENV MY_VARIABLE=

The MY_VARIABLE variable will be set on the container in Cycle as an empty string, but will be visible in interfaces such as the portal environment variable form as a reminder that the variable needs set.

Runtime Configuration Variables

Environment variables can be added to the container manually via the container runtime configuration. The variables do not have to exist on the image to be added this way.

Add runtime configuration variables

• Portal
• API

Scoped Variables

Scoped variables can be used when sharing variables between multiple different containers within an environment. Scoped variables can be injected into the container several different ways, one of which is as an environment variable.

Function Runtime Variables

Function containers can have environment variables injected dynamically on every start, so that each unique run of a function encapsulates its own environment.

Set Runtime Variables On Function Run

• API

Platform Environment Variables

The platform injects special "Cycle specific" environment variables into every container instance. This is to provide it with information about, for example, the environment its running in, the infrastructure provider, available resources, etc.

Order of Evaluation

When multiple environment variables of the same name are present from different sources, they're evaluated in the following order, where the last source wins.

1. Image Variables
2. Scoped Variables*
3. Runtime Configuration Variables
4. Function Runtime Variables
5. Cycle Platform Environment Variables

Environment Metadata File

In addition to the environment variables provided by the platform, Cycle also injects a JSON file into the container that contains information about the environment the container instance is running in. The file is mounted at:

/var/run/cycle/metadata/environment.json

This file is available at runtime, and may be useful for writing specialized scripts executed as an ENTRYPOINT into the container.

Example

{
	"id": "65b93b8dee5a5e8d87b2bd13",
	"deployments": {
		"tags": {
			"current": "v2.0.2",
			"demo": "v2.0.0",
			"live": "v8.9.10",
			"test": "v8.9.10"
		}
	},
	"private_network": {
		"vxlan_tag": 361,
		"subnet": "d600",
		"ipv6": {
			"ip": "fd00::d600:0:0:0",
			"cidr": "fd00::d600:0:0:0/80"
		},
		"legacy": null
	},
	"services": {
		"discovery": {
			"enable": true,
			"container_id": "65b93b8df9aadae178915a2e",
			"high_availability": false,
			"auto_update": false,
			"config": null
		},
		"vpn": {
			"enable": true,
			"container_id": "65b93b8df9aadae178915a32",
			"high_availability": false,
			"auto_update": false,
			"config": null
		},
		"loadbalancer": {
			"enable": true,
			"container_id": "65b93b8df9aadae178915a37",
			"high_availability": false,
			"auto_update": true,
			"config": null
		},
		"scheduler": {
			"enable": true,
			"container_id": "65b93b8df9aadae178915a39",
			"high_availability": false,
			"auto_update": false,
			"config": null
		}
	}
}