🚨 This is a work in progress 🚨
A terraform provider for authoring and executing tests using terraform primitives. Designed to work in conjunction with the Chainguard Images project. I would strongly recommend against using it for anything else.
See examples for usages, and design for more information about the providers design.
Basic acceptance tests:
IMAGETEST_ENTRYPOINT_REF=$(KO_DOCKER_REPO=ttl.sh/imagetest ko build ./cmd/entrypoint) \
TF_ACC=1 \
go test ./internal/provider/... -count=1 -v
This will build and use the entrypoint image, and use it in the test.
Testing the EKS driver takes a lot longer, and creates resources which might cost money. To run these tests, ensure you have AWS auth set up and eksctl installed, then run:
IMAGETEST_ENTRYPOINT_REF=$(KO_DOCKER_REPO=ttl.sh/imagetest ko build ./cmd/entrypoint) \
TF_ACC=1 \
go test -tags=eks ./internal/provider/... -count=1 -v -run=EKS -timeout=30m
This will run the EKS tests using your auth, and eksctl to manage the cluster.
The test will look like it's doing nothing for ~15-20 minutes, but you can check the progress by using eksctl get clusters.
When the cluster is up, you can find its kubeconfig in $TMPDIR/imagetest-<uid>, and eksctl populates the file (this takes ~10 minutes), use it to interact with the cluster:
KUBECONFIG=$TMPDIR/imagetest-<uid> kubectl get nodes
When the test completes, it should delete the cluster, but just in case it doesn't, you can delete it with:
eksctl delete cluster --name=imagetest-<uid>
You can also find the cluster in the AWS Console: https://us-west-2.console.aws.amazon.com/eks/home
To reuse the cluster instead of creating a new one each time, you can run the tests with IMAGETEST_EKS_SKIP_TEARDOWN=true.
Then, the next time you run the test, find the cluster that the last test created, and add IMAGETEST_EKS_CLUSTER=imagetest-<uid> to reuse the cluster.