Secrets in Jobs¶
Sometimes you want to use sensitive information within your code. Examples are: passwords, OAuth tokens, or ssh keys. The best practice for saving such information in Kubernetes is via Kubernetes Secrets. Kubernetes Secrets let you store and manage sensitive information. Access to secrets is limited via configuration.
A Kubernetes secret may hold multiple key - value pairs.
Using Secrets in Run:AI Jobs¶
Our goal is to provide Run:AI Jobs with secrets as an input in a secure way. Using the Run:AI command-line, you will be able to pass a reference to a secret that already exists in Kubernetes.
Creating a secret¶
For details on how to create a Kubernetes secret see: https://kubernetes.io/docs/concepts/configuration/secret/. Here is an example:
apiVersion: v1 kind: Secret metadata: name: my-secret namespace: runai-<project-name> data: username: am9obgo= password: bXktcGFzc3dvcmQK
kubectl apply -f <file-name>
- Secrets are base64 encoded
- Secrets are stored in the scope of a namespace and will not be accessible from other namespaces. Hence the reference to the Run:AI project name above. Run:AI provides the ability to propagate secrets throughout all Run:AI projects. See below.
Attaching a secret to a Job on Submit¶
When you start a job, you want to connect the secret to the new Job. To do that, run:
runai submit -e <ENV-VARIABLE>=SECRET:<secret-name>,<secret-key> ....
runai submit -i ubuntu -e MYUSERNAME=SECRET:my-secret,username
Secrets and Projects¶
As per the note above, secrets are namespace-specific. If your secret relates to all Run:AI projects, do the following to propagate the secret to all projects:
- When creating a secret, set the namespace to be
- Run the following once to allow Run:AI to propagate the secret:
kubectl label secret <secret_name> runai/cluster-wide="true" -n runai
Secrets and Templates¶
A Secret can be set at the template level. For additional information see template configuration