To run agents on-premise, Foundational provides container images that you can run in your environment and report data back to us via API.
To use that, you'll need a couple of pieces of information from our support team:
An agent identifier
A deploy key to our Github repository
Used to read
AWS credentials to our account in the form of access key ID and secret access key
This will allow us to communicate with the agent through an SQS queue
Setting up the container
The configuration for the container is passed through environment variables.
Depending on your setup, you'll need to define additional environment variagbles.
Check out the relevant connector's page here
Here's an example of running a basic agent:
docker run --restart unless-stopped -d -e FD_GITHUB_DEPLOY_KEY="xxxxxxxx" -e CREDS_ID=fd-supplied-identifier -e AWS_ACCESS_KEY_ID=abcdabcdabcdabcd -e AWS_SECRET_ACCESS_KEY=1234abcd1234abcd1234abcd public.ecr.aws/foundational-io/on-prem-agent-loader:latest
** replace specific environment variables with your own!
The containers support both arm64 and amd64 architectures.
Fast Failure Recovery
Our agent starts by setting up a few requirements. To make subsequent runs faster, you can map /app/on_prem_agent on the docker to a persistent volume. This makes resuming the agent a matter of 1-2 seconds.
To do that, add the -v flag in your docker execution command:
docker run -v <path_to_local_folder>:/app/on_prem_agent ...
Connector Setup
Visit the relevant connector's setup page to see how to configure it for on-premise access.
Full list here