tests: add smoketest using kind#4506
tests: add smoketest using kind#4506justinsb wants to merge 1 commit intoGoogleCloudPlatform:masterfrom
Conversation
c27223e to
9651ce8
Compare
|
This is obviously only a first step, but the reconciler did run: |
|
/assign @xiaoweim |
6136aa1 to
6f52a6b
Compare
cb795f6 to
488358e
Compare
.github/workflows/ci-presubmit.yaml
Outdated
| - uses: actions/setup-go@v5 | ||
| with: | ||
| go-version-file: 'go.mod' | ||
| - name: "Run scripts/github-actions/tests-scenarios-acquisition" |
There was a problem hiding this comment.
Out of curiosity, are we replacing the tests-scenarios-acquisition with the smoke test completely? The smoke test looks.
There was a problem hiding this comment.
Ah good point, I forgot about that problem.
So .github/workflows/ci-presubmit.yaml is # Generated by dev/tasks/generate-github-actions (line 15)
I think it was manually edited to add the additional tasks, and we do need to fix that. Two ways I can think of:
- Move the manual edits to a manually maintained yaml file (so they won't get deleted when we regenerate)
- Ensure that the automatic generation recreates these tasks by adding them to the locations that
dev/tasks/generate-github-actionslooks at
There was a problem hiding this comment.
Sounds good (and I like the second option for the script take care of the tasks) :)
|
/lgtm |
|
/assign @cheftako |
|
ping @cheftako |
3c0ef99 to
29d3890
Compare
|
/assign @xiaoweim |
|
Issue #5224 |
|
/lgtm |
The start of an e2e smoketest that deploys into a cluster with the operator. A few (bigger) todos: * We should figure out how to capture the GCP traffic * We should ensure that we are running the images we actually push
|
New changes are detected. LGTM label has been removed. |
|
[APPROVALNOTIFIER] This PR is NOT APPROVED This pull-request has been approved by: xiaoweim The full list of commands accepted by this bot can be found here. DetailsNeeds approval from an approver in each of these files:Approvers can indicate their approval by writing |
| timeout-minutes: 60 | ||
| steps: | ||
| - uses: actions/checkout@v4 | ||
| - uses: actions/setup-go@4dc6199c7b1a012772edbd06daecab0f50c9053c # v6.1.0 |
There was a problem hiding this comment.
Yuck. V6.2.0 is out. Could we use either the actual v6.1.0 or v6.2.0?
| @@ -0,0 +1,87 @@ | |||
| #!/usr/bin/env bash | |||
| # Copyright 2024 Google LLC | |||
| echo "Creating namespace ${NS}" | ||
| kubectl create ns ${NS} --dry-run=client -oyaml | kubectl apply --server-side -f - | ||
|
|
||
| # echo "Creating ConfigConnectorContext in namespace ${NS} (with fake google service account)" |
There was a problem hiding this comment.
Do we want these large commented out blocks? I would also suggest it is probably clearer to create a CCC function, a StorageBucket function and then have the main code path call (or not) those functions.
|
|
||
| if [[ -z "${VERSION:-}" ]]; then | ||
| echo "VERSION must be set" | ||
| if [[ -z "${IMAGE_TAG:-}" ]]; then |
There was a problem hiding this comment.
AGENT: Renaming VERSION to IMAGE_TAG is a breaking change for any existing callers of this script. Please check if this script is used elsewhere (e.g., local development workflows) or support both variables.
|
|
||
| kustomize build operator/config/default | \ | ||
| sed -e "s@image: operator:.*@image: gcr.io/gke-release/cnrm/operator:${VERSION}@g" | \ | ||
| sed -e "s@image: operator:.*@image: gcr.io/gke-release/cnrm/operator:${IMAGE_TAG}@g" | \ |
There was a problem hiding this comment.
TAKO: At some future point I think we should start plumbing in the registry root. ('gcr.io/gke-release'). Once that is done it becomes a lot easier to generate dev builds and run them.
| echo "Configuring namespace mode" | ||
| # # Configure in namespace mode, per instructions at https://cloud.google.com/config-connector/docs/how-to/install-namespaced | ||
| # echo "Configuring namespace mode" | ||
| # cat <<EOF | kubectl apply --server-side -f - |
There was a problem hiding this comment.
Still thinking making functions out of these and then have the mainline more like:
Setup-namespace-mode()
Setup_CC()
| # EOF | ||
|
|
||
| # Configure in cluster mode so we can use a (fake) google service account | ||
| echo "Creating namespace cnrm-system" |
There was a problem hiding this comment.
If we recommend namespace mode, then shouldn't out smoketest be to have namespace mode be the default test?
The start of an e2e smoketest that deploys into a cluster with the operator.
A few (bigger) todos: