Skip to content

Latest commit

 

History

History

krakend

README.md

Quick reference

Supported tags and respective Dockerfile links

Quick reference (cont.)

logo

What is KrakenD?

KrakenD is a stateless, high-performance, enterprise-ready, open-source API gateway written in Go. Its engine (formerly known as KrakenD Framework) is now a Linux Foundation Project codenamed Lura Project. Lura is the only enterprise-grade API Gateway hosted in a neutral, open forum.

KrakenD is lightweight and straightforward, as it only requires writing the configuration file. No Go knowledge is required. It offers connectivity to internal and external services, data transformation and filtering, and aggregation of multiple data sources (APIs, gRPC, queues and pub/sub, lambda, etc.) simultaneously or in cascade. It protects access to your API, throughputs its usage, and integrates with many third-parties.

All features are designed to offer extraordinary performance and infinite scalability.

How to use this image

KrakenD only needs a single configuration file to create an API Gateway, although you can have a complex setup reflecting your organization structure. The configuration file(s) can live anywhere in the container, but the default location (the workdir) is /etc/krakend.

To use the image, COPY your krakend.json file inside the container or mount it using a volume. The configuration is checked only once during the startup and never used again. Don't have a config file yet? Generate it with the KrakenD Designer UI.

⚠️ NOTICE: KrakenD does not use live reload when your configuration changes. Restart the container.

Quick start

You can start an empty gateway with a health check with the following commands:

docker run -d -p 8080:8080 -v "$PWD:/etc/krakend/" krakend

curl http://localhost:8080/__health
{"agents":{},"now":"2024-05-23 14:35:55.552591448 +0000 UTC m=+26.856583003","status":"ok"}

More Examples

The following are several examples of running KrakenD. By default, the command run is executed, but you can pass other commands and flags at the end of the run command.

The configuration files are taken from the current directory ($PWD). Therefore, all examples expect to find at least the file krakend.json.

Run with the debug enabled (flag -d):

This flag is SAFE to use in production. It's meant to enable KrakenD as a fake backend itself by enabling a /__debug endpoint

docker run -p 8080:8080 -v "${PWD}:/etc/krakend/" krakend run -d -c /etc/krakend/krakend.json

Checking the syntax of your configuration file

See the check command

docker run -it -v $PWD:/etc/krakend/ krakend check --config krakend.json

Show the help:

docker run --rm -it krakend help

Building your custom KrakenD image

Most production deployments will not want to rely on mounting a volume for the container but to use their image based on krakend:

Your Dockerfile could look like this:

FROM krakend:<version>
# NOTE: Avoid using :latest image on production. Stick to a major version instead.

COPY krakend.json ./

# Check and test that the file is valid
RUN krakend check -t --lint-no-network -c krakend.json

If you want to manage your KrakenD configuration using multiple files and folders, reusing templates, and distributing the configuration amongst your teams, you can use the flexible configuration (FC). The following Dockerfile combines FC, the krakend check command, and a 2-step build.

FROM krakend:<version> as builder

COPY krakend.tmpl .
COPY config .

# Save temporary output file to /tmp to avoid permission errors
RUN FC_ENABLE=1 \
    FC_OUT=/tmp/krakend.json \
    FC_PARTIALS="/etc/krakend/partials" \
    FC_SETTINGS="/etc/krakend/settings" \
    FC_TEMPLATES="/etc/krakend/templates" \
    krakend check -d -t -c krakend.tmpl

# Copy the output file only and discard any other files
FROM krakend:<version>
COPY --from=builder /tmp/krakend.json .

Then build with docker build -t my_krakend .

The configuration above assumes you have a folder structure like the following:

.
├── config
│   ├── partials
│   ├── settings
│   │   └── env.json
│   └── templates
│       └── some.tmpl
├── Dockerfile
└── krakend.tmpl

Docker Compose example

Finally, a simple docker-compose file to start KrakenD with your API would be:

version: "3"
services:
  krakend:
    image: krakend:<version>
    ports:
      - "8080:8080"
    volumes:
      - ./:/etc/krakend

And another one that uses the flexible configuration and a custom template filename (my_krakend.tmpl) on each start:

version: "3"
services:
  krakend:
    image: krakend:<version>
    ports:
      - "8080:8080"
    volumes:
      - ./:/etc/krakend
    environment:
      - FC_ENABLE=1
      - FC_OUT=/tmp/krakend.json
      - FC_PARTIALS="/etc/krakend/config/partials"
      - FC_SETTINGS="/etc/krakend/config/settings/prod"
      - FC_TEMPLATES="/etc/krakend/config/templates"
    command:
      command: ["krakend", "run", "-c", "krakend.tmpl", "-d"]

Container permissions and commands

All krakend commands are executed as krakend user (uid=1000), and the rest of the commands (e.g., sh) are executed as root.

You can directly use sub-commands of krakend like run, help, version, check, check-plugin, or test-plugin as the entrypoint will add the krakend command automatically. For example, the following lines are equivalent:

docker run --rm -it krakend help
docker run --rm -it krakend krakend help

License

View license information for the software contained in this image.

As with all Docker images, these likely also contain other software which may be under other licenses (such as Bash, etc from the base distribution, along with any direct or indirect dependencies of the primary software being contained).

Some additional license information which was able to be auto-detected might be found in the repo-info repository's krakend/ directory.

As for any pre-built image usage, it is the image user's responsibility to ensure that any use of this image complies with any relevant licenses for all software contained within.