Skip to content

Add Dockerfile, Ankra application descriptor, Kubernetes manifests and CI/CD pipeline - #1

Open
ankra-platform[bot] wants to merge 3 commits into
mainfrom
ankra/setup-tael-cli
Open

ankra-platform[bot] wants to merge 3 commits into
mainfrom
ankra/setup-tael-cli

Conversation

@ankra-platform

Copy link
Copy Markdown

Ankra Application Setup

Detected stack: go

This PR adds the following generated files:

  • Dockerfile
  • .ankra/ankra.yaml
  • .ankra/manifests/deployment.yaml
  • .ankra/manifests/service.yaml
  • .ankra/manifests/ingress.yaml
  • .github/workflows/build-and-publish.yml

What's included

  • Dockerfile - multi-stage build optimized for the detected stack
  • GitHub Actions workflow - builds and pushes the container image to the private Ankra registry
  • Kubernetes manifests - 3 file(s), one per resource, under .ankra/manifests/, parameterised with Ankra inputs and rendered per environment
  • HTTPS Ingress, switched off until you name a host - set ingress_enabled and ingress_host on the deploy form to publish it. ingress_tls_issuer names the cert-manager ClusterIssuer that mints the certificate (letsencrypt-prod on Ankra-provisioned clusters); clear it to serve plain HTTP.
  • Health probes check the container port, not a URL - Ankra found no health route in the source it read, and probing a path your application does not serve would keep the pod from ever becoming ready. A TCP connection to the port answers the narrower question truthfully. Commit a health route and re-run setup to have the probes use it.
  • Ankra application descriptor - .ankra/ankra.yaml, this application's metadata and its deploy option contract; Ankra reads it back on every analysis run, so edit the options here to change the deploy form
  • Values your application still needs - your code reads these and nothing generated here supplies them: TAEL_CONFIG. Set them on this application's Environment secrets panel in Ankra; the next deploy seals them into the tael-cli-env Secret in the target namespace. Nothing is committed to this repository and nothing has to be created by hand. Values the platform provides - the generated database's credentials among them - are already wired and are deliberately not listed here.

Review the generated files and merge when ready. Once merged, Ankra deploys these manifests to your clusters as an application in a stack.

Generated by Ankra

ankra-platform Bot and others added 3 commits August 26, 2026 15:02
…ests and ci/cd pipeline

Generated by Ankra platform
The generated manifests modelled a listening HTTP service: a Service on
port 80, an Ingress, TCP probes against container port 8080, and a
target_port option. The CLI binds no listener, so all of that described
something that does not exist.

The pod now idles signal-responsively so operators can exec in and run
tael commands against the platform from inside the cluster; health is an
exec of 'tael version', the only meaningful signal for a toolbox. Adds a
tael_base_url option and a writable HOME for CLI config.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01JMj6ZLzueMmZMDMWSLmQGy
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant