Running Keycloak in a container

Learn how to run Keycloak from a container image

This guide describes how to optimize and run the Keycloak container image to provide the best experience running a container.

Creating a customized and optimized container image

The default Keycloak container image ships ready to be configured and optimized.

For the best start up of your Keycloak container, build an image by running the build step during the container build. This step will save time in every subsequent start phase of the container image.

Writing your optimized Keycloak Dockerfile

The following Dockerfile creates a pre-configured Keycloak image that enables the health and metrics endpoints, enables the token exchange feature, and uses a PostgreSQL database.

FROM as builder

# Enable health and metrics support

# Configure a database vendor
ENV KC_DB=postgres

WORKDIR /opt/keycloak
# for demonstration purposes only, please make sure to use proper certificates in production instead
RUN keytool -genkeypair -storepass password -storetype PKCS12 -keyalg RSA -keysize 2048 -dname "CN=server" -alias server -ext "SAN:c=DNS:localhost,IP:" -keystore conf/server.keystore
RUN /opt/keycloak/bin/ build

COPY --from=builder /opt/keycloak/ /opt/keycloak/

# change these values to point to a running postgres instance
ENV KC_DB=postgres
ENTRYPOINT ["/opt/keycloak/bin/"]

The build process includes multiple stages:

  • Run the build command to set server build options to create an optimized image.

  • The files generated by the build stage are copied into a new image.

  • In the final image, additional configuration options for the hostname and database are set so that you don’t need to set them again when running the container.

  • In the entrypoint, the enables access to all the distribution sub-commands.

To install custom providers, you just need to define a step to include the JAR file(s) into the /opt/keycloak/providers directory. This step must be placed before the line that RUNs the build command, as below:

# A example build step that downloads a JAR file from a URL and adds it to the providers directory
FROM as builder


# Add the provider JAR file to the providers directory
ADD --chown=keycloak:keycloak --chmod=644 <MY_PROVIDER_JAR_URL> /opt/keycloak/providers/myprovider.jar


# Context: RUN the build command
RUN /opt/keycloak/bin/ build

Installing additional RPM packages

If you try to install new software in a stage FROM, you will notice that microdnf, dnf, and even rpm are not installed. Also, very few packages are available, only enough for a bash shell, and to run Keycloak itself. This is due to security hardening measures, which reduce the attack surface of the Keycloak container.

First, consider if your use case can be implemented in a different way, and so avoid installing new RPMs into the final container:

  • A RUN curl instruction in your Dockerfile can be replaced with ADD, since that instruction natively supports remote URLs.

  • Some common CLI tools can be replaced by creative use of the Linux filesystem. For example, ip addr show tap0 becomes cat /sys/class/net/tap0/address

  • Tasks that need RPMs can be moved to a former stage of an image build, and the results copied across instead.

Here is an example. Running update-ca-trust in a former build stage, then copying the result forward:

FROM AS ubi-micro-build
COPY mycertificate.crt /etc/pki/ca-trust/source/anchors/mycertificate.crt
RUN update-ca-trust

COPY --from=ubi-micro-build /etc/pki /etc/pki

It is possible to install new RPMs if absolutely required, following this two-stage pattern established by ubi-micro:

FROM AS ubi-micro-build
RUN mkdir -p /mnt/rootfs
RUN dnf install --installroot /mnt/rootfs <package names go here> --releasever 9 --setopt install_weak_deps=false --nodocs -y && \
    dnf --installroot /mnt/rootfs clean all && \
    rpm --root /mnt/rootfs -e --nodeps setup

COPY --from=ubi-micro-build /mnt/rootfs /

This approach uses a chroot, /mnt/rootfs, so that only the packages you specify and their dependencies are installed, and so can be easily copied into the second stage without guesswork.

Some packages have a large tree of dependencies. By installing new RPMs you may unintentionally increase the container’s attack surface. Check the list of installed packages carefully.

Building the container image

To build the actual container image, run the following command from the directory containing your Dockerfile:

podman|docker build . -t mykeycloak

Starting the optimized Keycloak container image

To start the image, run:

podman|docker run --name mykeycloak -p 8443:8443 -p 9000:9000 \
        -e KEYCLOAK_ADMIN=admin -e KEYCLOAK_ADMIN_PASSWORD=change_me \
        mykeycloak \
        start --optimized

Keycloak starts in production mode, using only secured HTTPS communication, and is available on https://localhost:8443.

Opening up https://localhost:9000/metrics leads to a page containing operational metrics that could be used by your monitoring solution.

Exposing the container to a different port

By default, the server is listening for http and https requests using the ports 8080 and 8443, respectively.

If you want to expose the container using a different port, you need to set the hostname-port accordingly:

  1. Exposing the container using a port other than the default ports

podman|docker run --name mykeycloak -p 3000:8443 \
        -e KEYCLOAK_ADMIN=admin -e KEYCLOAK_ADMIN_PASSWORD=change_me \
        mykeycloak \
        start --optimized --hostname-port=3000

By setting the hostname-port option you can now access the server at https://localhost:3000.

Trying Keycloak in development mode

The easiest way to try Keycloak from a container for development or testing purposes is to use the Development mode. You use the start-dev command:

podman|docker run --name mykeycloak -p 8080:8080 \
        -e KEYCLOAK_ADMIN=admin -e KEYCLOAK_ADMIN_PASSWORD=change_me \ \

Invoking this command starts the Keycloak server in development mode.

This mode should be strictly avoided in production environments because it has insecure defaults. For more information about running Keycloak in production, see Configuring Keycloak for production.

Running a standard Keycloak container

In keeping with concepts such as immutable infrastructure, containers need to be re-provisioned routinely. In these environments, you need containers that start fast, therefore you need to create an optimized image as described in the preceding section. However, if your environment has different requirements, you can run a standard Keycloak image by just running the start command. For example:

podman|docker run --name mykeycloak -p 8080:8080 \
        -e KEYCLOAK_ADMIN=admin -e KEYCLOAK_ADMIN_PASSWORD=change_me \ \
        start \
        --db=postgres --features=token-exchange \
        --db-url=<JDBC-URL> --db-username=<DB-USER> --db-password=<DB-PASSWORD> \
        --https-key-store-file=<file> --https-key-store-password=<password>

Running this command starts a Keycloak server that detects and applies the build options first. In the example, the line --db=postgres --features=token-exchange sets the database vendor to PostgreSQL and enables the token exchange feature.

Keycloak then starts up and applies the configuration for the specific environment. This approach significantly increases startup time and creates an image that is mutable, which is not the best practice.

Provide initial admin credentials when running in a container

Keycloak only allows to create the initial admin user from a local network connection. This is not the case when running in a container, so you have to provide the following environment variables when you run the image:

# setting the admin username
-e KEYCLOAK_ADMIN=<admin-user-name>

# setting the initial password

Importing A Realm On Startup

The Keycloak containers have a directory /opt/keycloak/data/import. If you put one or more import files in that directory via a volume mount or other means and add the startup argument --import-realm, the Keycloak container will import that data on startup! This may only make sense to do in Dev mode.

podman|docker run --name keycloak_unoptimized -p 8080:8080 \
        -e KEYCLOAK_ADMIN=admin -e KEYCLOAK_ADMIN_PASSWORD=change_me \
        -v /path/to/realm/data:/opt/keycloak/data/import \ \
        start-dev --import-realm

Feel free to join the open GitHub Discussion around enhancements of the admin bootstrapping process.

Specifying different memory settings

The Keycloak container, instead of specifying hardcoded values for the initial and maximum heap size, uses relative values to the total memory of a container. This behavior is achieved by JVM options -XX:MaxRAMPercentage=70, and -XX:InitialRAMPercentage=50.

The -XX:MaxRAMPercentage option represents the maximum heap size as 70% of the total container memory. The -XX:InitialRAMPercentage option represents the initial heap size as 50% of the total container memory. These values were chosen based on a deeper analysis of Keycloak memory management.

As the heap size is dynamically calculated based on the total container memory, you should always set the memory limit for the container. Previously, the maximum heap size was set to 512 MB, and in order to approach similar values, you should set the memory limit to at least 750 MB. For smaller production-ready deployments, the recommended memory limit is 2 GB.

The JVM options related to the heap might be overridden by setting the environment variable JAVA_OPTS_KC_HEAP. You can find the default values of the JAVA_OPTS_KC_HEAP in the source code of the, or kc.bat script.

For example, you can specify the environment variable and memory limit as follows:

podman|docker run --name mykeycloak -p 8080:8080 -m 1g \
        -e KEYCLOAK_ADMIN=admin -e KEYCLOAK_ADMIN_PASSWORD=change_me \
        -e JAVA_OPTS_KC_HEAP="-XX:MaxHeapFreeRatio=30 -XX:MaxRAMPercentage=65" \ \
If the memory limit is not set, the memory consumption rapidly increases as the heap size can grow up to 70% of the total container memory. Once the JVM allocates the memory, it is returned to the OS reluctantly with the current Keycloak GC settings.

Relevant options



The database vendor.

CLI: --db
Env: KC_DB

dev-file (default), dev-mem, mariadb, mssql, mysql, oracle, postgres


The password of the database user.

CLI: --db-password


The full database JDBC URL.

If not provided, a default URL is set based on the selected database vendor. For instance, if using postgres, the default JDBC URL would be jdbc:postgresql://localhost/keycloak.

CLI: --db-url


The username of the database user.

CLI: --db-username


Enables a set of one or more features.

CLI: --features

account-api[:v1], account3[:v1], admin-api[:v1], admin-fine-grained-authz[:v1], admin2[:v1], authorization[:v1], ciba[:v1], client-policies[:v1], client-secret-rotation[:v1], client-types[:v1], declarative-ui[:v1], device-flow[:v1], docker[:v1], dpop[:v1], dynamic-scopes[:v1], fips[:v1], hostname[:v1,v2], impersonation[:v1], js-adapter[:v1], kerberos[:v1], login2[:v1], multi-site[:v1], oid4vc-vci[:v1], organization[:v1], par[:v1], passkeys[:v1], persistent-user-sessions[:v1], preview, recovery-codes[:v1], scripts[:v1], step-up-authentication[:v1], token-exchange[:v1], transient-users[:v1], update-email[:v1], web-authn[:v1]


Address at which is the server exposed.

Can be a full URL, or just a hostname. When only hostname is provided, scheme, port and context path are resolved from the request.

CLI: --hostname

Available only when hostname:v2 feature is enabled


Hostname for the Keycloak server.

CLI: --hostname

Available only when hostname:v1 feature is enabled



The key store which holds the certificate information instead of specifying separate files.

CLI: --https-key-store-file


The password of the key store file.

CLI: --https-key-store-password

password (default)


If the server should expose health check endpoints.

If enabled, health checks are available at the /health, /health/ready and /health/live endpoints.

CLI: --health-enabled

true, false (default)


If the server should expose metrics.

If enabled, metrics are available at the /metrics endpoint.

CLI: --metrics-enabled

true, false (default)

On this page