Skip to main content

Docker Entrypoint

docker-entrypoint is a basic utility to proxy a container's normal entrypoint. This avoid a few pitfalls in developing and debugging processes running in kubernetes pods, namely:

  • Starting/stopping debugging sessions without restarting the container, and subsequently losing state of the filesystem.
  • Changing environment variables without restarting the container, so environment changes can be made without redeploying to kubernetes.

How to Install

docker-entrypoint can be installed in your docker image via pip:

pip install docker-entrypoint

Since there is no mechanism for injecting the entrypoint script into your image, you'll likely want to install docker-entrypoint at image build time (i.e. in your Dockerfile). Here's a basic, incomplete example:

FROM python:3.7
...
# Install via pip:
RUN pip install docker-entrypoint

# Continue as normal:
WORKDIR /path/to/my/app
ENTRYPOINT ["python", "main.py"]

How to Use

Running your process through entrypoint

docker-entrypoint will install an entrypoint driver globally which accepts your normal entrypoint command as arguments.

root@my-pod:/path/to/my/app# entrypoint python main.py

Since entrypoint must be running as PID 1 in the pod's container to have any value, we must modify the configured entrypoint for the container at launch. There are two options to do so.

Option one: Kubernetes Container Spec

Specify an alternate entrypoint in the container spec for your pod/deploy/statefulset manifest:

spec:
  containers:
  - name: my-app
    command: ["entrypoint"]
    args: ["python", "main.py"]

Option two: Dockerfile ENTRYPOINT

Use entrypoint as your entrypoint in the Dockerfile itself:

FROM python:3.7
...
# Install via pip:
RUN pip install docker-entrypoint

# Note the ENTRYPOINT change:
WORKDIR /path/to/my/app
ENTRYPOINT ["entrypoint", "python", "main.py"]

Starting a debug session

Simply import the Debugger class, make an instance, and call Debugger.start()

from entrypoint.debug import Debugger

dbg = Debugger()
dbg.start() # Will wait for debugger to attach, depending on configuration.

docker-entrypoint includes a module that supports remote debugging for VS Code and PyCharm. The Debugger class currently makes several assumptions:

Environment Variable Valid Options Default
DEBUGGER_ENABLED true, false false
DEBUGGER_ADDRESS any hostname/IP localhost
DEBUGGER_EDITOR vscode, pycharm vscode
DEBUGGER_WAIT true, false true

These values are pulled from a "dotenv" file located at /tmp/.env. This is hardcoded. Additionally, the ports used for forwarding and reverse forwarding are hardcoded to 9000 and 9001, respectively. When launching port-forwarding commands, be sure to use these ports for the pod-side binding.

Get the name of your pod:

POD_NAME=$(kubectl get pods -l release=my-app --field-selector status.phase=Running -o jsonpath={.items[0].metadata.name})

Restart your process. This does not restart your container, only the process being managed by entrypoint.

kubectl exec $POD_NAME -- bash -c "kill -HUP 1"

Environment variables are loaded from the in-pod .env file prior to each launch of the debugger, so they can be updated on-demand from a local .env file. Simply copy your local .env file into the pod and restart your process. Again, this does not restart the container.

kubectl cp .env $POD_NAME:/tmp/.env
kubectl exec $POD_NAME -- bash -c "kill -HUP 1"

For IDEs that create a debug server and require the remote process to attach to it (i.e. PyCharm), Debugger.start() will wait until SIGUSR1 OS signal is received. Start your debug server in your IDE and send your pod a signal when its waiting for a connection.

kubectl exec $POD_NAME -- bash -c "kill -USR1 1"

Module Breakdown

main.py

Defines an asyncio loop for managing the process proxied by entrypoint. This includes forwarding relevant signals and restarting your process when it exits. The following table documents the signals that are currently handled:

Signal Outcome
SIGTERM Causes entrypoint to stop managing and exit. This restarts the container as normal.
SIGINT Causes entrypoint to stop managing and exit. This restarts the container as normal.
SIGHUP Signal entrypoint to forward SIGTERM to process, and then spawns another process.
SIGUSR1 Signal entrypoint to continue with establishing a debugging session. Only necessary when DEBUGGER_EDITOR=pycharm.

debug.py

Defines a Debugger class that will start a debugging session based on environment variables (as documented here).

Debugger.__init__()

Loads environment variables into os.environ from a dotenv file located at /tmp/.env, and then parses these values for use in a debugging session.

Debugger.start()

Starts a debugging session.

Metadata

Release files for docker-entrypoint 0.1.7

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for docker-entrypoint 0.1.7
File Size Uploaded Provenance Actions
docker_entrypoint-0.1.7.tar.gz 11.7 kB With provenance Details for docker_entrypoint-0.1.7.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for docker-entrypoint 0.1.7
File Interpreter ABI Platform Actions
docker_entrypoint-0.1.7-py3-none-any.whl Python 3 none any Details for docker_entrypoint-0.1.7-py3-none-any.whl

Total release size: 23.8 kB

Release files / docker_entrypoint-0.1.7.tar.gz

Download URL docker_entrypoint-0.1.7.tar.gz
Size 11.7 kB
Tags Source
SHA-256 checksum
How to use checksums
a356f30f35f3de7b7f0b8d7345e13bc5ef1a53310f8185c35cf54e83f544a30b
BLAKE2b-256 checksum
How to use checksums
9e7b449ce08d94bc337609e1b5ab4622527eab9a7eb6f8ffb3182fe66faf6e13
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Sep 8, 2026.

Transparency log

Release files / docker_entrypoint-0.1.7-py3-none-any.whl

Download URL docker_entrypoint-0.1.7-py3-none-any.whl
Size 12.1 kB
Tags Python 3
SHA-256 checksum
How to use checksums
bf55190fa23fbb5e5bd1831084bae22987430f168fe79157422a7f91cf0c940f
BLAKE2b-256 checksum
How to use checksums
2125d7c39de073ce3b9d64e4b68d2a72ac07c6a152605de20eb714f1ab700c40
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Sep 8, 2026.

Transparency log

Security for docker-entrypoint 0.1.7

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

Full provenance

All 2 files in this release have been uploaded with provenance metadata.

Select from 2 files

Attestation for docker_entrypoint-0.1.7.tar.gz

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Sep 8, 2026.

Transparency log

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page