You can create custom flows for a template, that allow you to run whatever you want (bash, python, gcloud, ansible, cloudformation, etc.), whenever you want in the deployment process (before or after Terraform init/plan/apply, and even destroy/error).
Create a file named env0.yml to define a custom flow. env0 will use this file it is defined.
Below is an example of a custom flow which creates a more "dynamic" Terraform file, based on an environment variable value, and uses Python and Jinja before the Terraform Init (the full example of this template git repo (including the env0.yml file) is here):
version: 1
deploy:
steps:
terraformInit:
before:
- pip install --user j2cli
- 'echo Generating "stage.auto.tfvars" for "$STAGE"'
- j2 stage.auto.tfvars.j2 > stage.auto.tfvars
- 'echo -e "Generated \"stage.auto.tfvars\": \n$(cat stage.auto.tfvars)"'
- 'echo Generating "elb.tf" for "$STAGE"'
- j2 elb.tf.j2 > elb.tf
- 'echo -e "Generated \"elb.tf\": \n$(cat elb.tf)"'
You can include the following possible hooks in an env0.yml file:
version:
type: number
ā
deploy:
onCompletion:
- type: string
onSuccess:
- type: string
onFailure:
- type: string
ā
steps:
setupVariables:
after:
- type: string
ā
terraformInit:
before:
- type: string
after:
- type: string
ā
terraformPlan:
before:
- type: string
after:
- type: string
ā
terraformApply:
before:
- type: string
after:
- type: string
ā
storeState:
before:
- type: string
after:
- type: string
ā
terraformOutput:
before:
- type: string
after:
- type: string
ā
destroy:
onCompletion:
- type: string
onSuccess:
- type: string
onFailure:
- type: string
ā
steps:
setupVariables:
after:
- type: string
ā
terraformInit:
before:
- type: string
after:
- type: string
ā
terraformPlan:
before:
- type: string
after:
- type: string
ā
terraformDestroy:
before:
- type: string
after:
- type: string
Handling errors
If any command in your custom flow returns a non-zero exit code, the env0 deployment will stop on a 'Failed' status. You can leverage this to perform validation - for example - asserting a certain environment variable was supplied.
In such a case, env0 will display any output from stderr as the deployment error. If no output is found in stderr, it will be shown as an "Unknown Error".
Here is an example -
version:
type: 1
deploy:
steps:
setupVariables:
after:
- if [ -z "$MY_VAR" ]; then echo "MY_VAR must be supplied" 1>&2 && exit 1; fi
The Deployment Container
Our deployment container is based on node:12-alpine and includes the following:
Includes | Version |
|---|---|
Bash | 5.0.11(1)-release |
NodeJS | 12.18.3 |
NPM | 6.14.6 |
YARN | 1.22.4 |
GIT | 2.24.3 |
Python2 | 2.7.18 |
Python3 | 3.8.2 |
pip3 | 20.2 |
pip2 | 18.1 |
AWS CLI | 1.18.115 |
Google Cloud SDK | 304.0.0 |
Azure CLI | 2.10.1 |
kubectl | 1.15.11 |
tfenv | 1.0.1 |
Terraform default version | 0.12.24 |
jq | jq-master-v20191114-85-g260888d269 |
OpenSSL | 1.1.1g |
Powershell | 7.0.3 |
Updated 12 days ago