entrypoint

README

entrypoint

This binary is used to override the entrypoint of a container by wrapping it and executing original entrypoint command in a subprocess.

Tekton uses this to make sure TaskRuns' steps are executed in order, only after sidecars are ready and previous steps have completed successfully.

Flags

The following flags are available:

• -entrypoint: "original" command to be executed (as entrypoint). This will be executed as a sub-process on entrypoint
• -post_file: file path to write once the sub-process has finished. If the sub-process failed, it will write to {{post_file}}.err instead of {{post_file}}.
• -wait_file: file path to watch before starting the sub-process. It watches for {{wait_file}} and {{wait_file}}.err presence and will either execute the sub-process (in case of {{wait_file}}) or skip the execution, write to {{post_file}}.err and return an error (exitCode >= 0)
• -wait_file_content: expects the wait_file to contain actual contents. It will continue watching for wait_file until it has content.

Any extra positional arguments are passed to the original entrypoint command.

Example

The following example of usage for entrypoint waits for /tekton/tools/3 file to exist and executes the command bash with args echo and hello, then writes the file /tekton/tools/4, or /tekton/tools/4.err in case the command fails.

entrypoint \
  -wait_file /tekton/tools/3 \
  -post_file /tekton/tools/4 \
  -entrypoint bash -- \
  echo hello

Waiting for Sidecars

In cases where the TaskRun's Pod has sidecar containers -- including, possibly, injected sidecars that Tekton itself didn't specify -- the first step should also wait until all those sidecars have reported as ready. Starting before sidecars are ready could lead to flaky errors if steps rely on the sidecar being ready to succeed.

To account for this, the Tekton controller starts TaskRun Pods with the first step's entrypoint binary configured to wait for a special file provided by the Kubernetes Downward API. This allows Tekton to write a Pod annotation when all sidecars report as ready, and for the value of that annotation to appear to the Pod as a file in a Volume. To the Pod, that file always exists, but without content until the annotation is set, so we instruct the entrypoint to wait for the -wait_file to contain contents before proceeding.

Example

The following example of usage for entrypoint waits for /tekton/downward/ready file to exist and contain actual contents (-wait_file_contents), and executes the command bash with args echo and hello, then writes the file /tekton/tools/1, or /tekton/tools/1.err in case the command fails.

entrypoint \
  -wait_file /tekton/downward/ready \
  -wait_file_contents \
  -post_file /tekton/tools/1 \
  -entrypoint bash -- \
  echo hello

cp Mode

In order to make the entrypoint binary available to the user's steps, it gets copied to a Volume that's shared with all the steps' containers. This is done in an initContainer pre-step, that runs before steps start.

To reduce external dependencies, the entrypoint binary actually copies itself to the shared Volume. When executed with the positional args of cp <src> <dst>, the entrypoint binary copies the <src> file to <dst> and exits.

It's executed as an initContainer in the TaskRun's Pod like:

initContainers:
- image: gcr.io/tekton-releases/github.com/tektoncd/pipeline/cmd/entrypoint
  args:
  - cp
  - /ko-app/entrypoint  # <-- path to the entrypoint binary inside the image
  - /tekton/tools/entrypoint
  volumeMounts:
  - name: tekton-internal-tools
    mountPath: /tekton/tools

containers:
- image: user-image
  command:
  - /tekton/tools/entrypoint
  ... args to entrypoint ...
  volumeMounts:
  - name: tekton-internal-tools
    mountPath: /tekton/tools

volumes:
- name: tekton-internal-tools
  volumeSource:
    emptyDir: {}