mirror of
https://github.com/keycloak/keycloak.git
synced 2026-01-09 23:12:06 -03:30
497 lines
18 KiB
Plaintext
497 lines
18 KiB
Plaintext
<#import "/templates/guide.adoc" as tmpl>
|
|
<#import "/templates/kc.adoc" as kc>
|
|
<#import "/templates/options.adoc" as opts>
|
|
<#import "/templates/links.adoc" as links>
|
|
<#import "/templates/profile.adoc" as profile>
|
|
|
|
<@tmpl.guide
|
|
title="Advanced configuration"
|
|
summary="How to tune advanced aspects of the Keycloak CR">
|
|
|
|
== Advanced configuration
|
|
This {section} describes how to use Custom Resources (CRs) for advanced configuration of your {project_name} deployment.
|
|
|
|
=== Server configuration details
|
|
|
|
Many server options are exposed as first-class citizen fields in the Keycloak CR. The structure of the CR is based on the configuration structure of {project_name}. For example, to configure the `https-port` of the server, follow a
|
|
similar pattern in the CR and use the `httpsPort` field. The following example is a complex server configuration; however, it illustrates the relationship between server options and the Keycloak CR:
|
|
|
|
[source,yaml]
|
|
----
|
|
apiVersion: k8s.keycloak.org/v2alpha1
|
|
kind: Keycloak
|
|
metadata:
|
|
name: example-kc
|
|
spec:
|
|
db:
|
|
vendor: postgres
|
|
usernameSecret:
|
|
name: usernameSecret
|
|
key: usernameSecretKey
|
|
passwordSecret:
|
|
name: passwordSecret
|
|
key: passwordSecretKey
|
|
host: host
|
|
database: database
|
|
port: 123
|
|
schema: schema
|
|
poolInitialSize: 1
|
|
poolMinSize: 2
|
|
poolMaxSize: 3
|
|
http:
|
|
httpEnabled: true
|
|
httpPort: 8180
|
|
httpsPort: 8543
|
|
tlsSecret: my-tls-secret
|
|
hostname:
|
|
hostname: https://my-hostname.tld
|
|
admin: https://my-hostname.tld/admin
|
|
strict: false
|
|
backchannelDynamic: true
|
|
features:
|
|
enabled:
|
|
- docker
|
|
- authorization
|
|
disabled:
|
|
- admin
|
|
- step-up-authentication
|
|
transaction:
|
|
xaEnabled: false
|
|
----
|
|
|
|
For a list of options, see the Keycloak CRD. For details on configuring options, see <@links.server id="all-config"/>.
|
|
|
|
==== Additional options
|
|
|
|
Some expert server options are unavailable as dedicated fields in the Keycloak CR. The following are examples of omitted fields:
|
|
|
|
* Fields that require deep understanding of the underlying {project_name} implementation
|
|
* Fields that are not relevant to
|
|
<@profile.ifCommunity>
|
|
a Kubernetes
|
|
</@profile.ifCommunity>
|
|
<@profile.ifProduct>
|
|
an OpenShift
|
|
</@profile.ifProduct>
|
|
environment
|
|
* Fields for provider configuration because they are dynamic based on the used provider implementation
|
|
|
|
The `additionalOptions` field of the Keycloak CR enables {project_name} to accept any available configuration in the form of key-value pairs.
|
|
You can use this field to include any option that is omitted in the Keycloak CR.
|
|
For details on configuring options, see <@links.server id="all-config"/>.
|
|
|
|
The values can be expressed as plain text strings or Secret object references as shown in this example:
|
|
|
|
[source,yaml]
|
|
----
|
|
apiVersion: k8s.keycloak.org/v2alpha1
|
|
kind: Keycloak
|
|
metadata:
|
|
name: example-kc
|
|
spec:
|
|
...
|
|
additionalOptions:
|
|
- name: spi-connections-http-client-default-connection-pool-size
|
|
secret: # Secret reference
|
|
name: http-client-secret # name of the Secret
|
|
key: poolSize # name of the Key in the Secret
|
|
- name: spi-email-template-mycustomprovider-enabled
|
|
value: true # plain text value
|
|
----
|
|
NOTE: The name format of options defined in this way is identical to the key format of options specified in the configuration file.
|
|
For details on various configuration formats, see <@links.server id="configuration"/>.
|
|
|
|
=== Secret References
|
|
|
|
Secret References are used by some dedicated options in the Keycloak CR, such as `tlsSecret`, or as a value in `additionalOptions`.
|
|
|
|
Similarly ConfigMap References are used by options such as the `configMapFile`.
|
|
|
|
When specifying a Secret or ConfigMap Reference, make sure that a Secret or ConfigMap containing the referenced keys is present in the same namespace as the CR referencing it.
|
|
|
|
The operator will poll approximately every minute for changes to referenced Secrets or ConfigMaps. When a meaningful change is detected, the Operator performs a rolling restart of the {project_name} Deployment to pick up the changes.
|
|
|
|
=== Unsupported features
|
|
|
|
The `unsupported` field of the CR contains highly experimental configuration options that are not completely tested and are Tech Preview.
|
|
|
|
==== Pod Template
|
|
|
|
The Pod Template is a raw API representation that is used for the Deployment Template.
|
|
This field is a temporary workaround in case no supported field exists at the top level of the CR for your use case.
|
|
|
|
The Operator merges the fields of the provided template with the values generated by the Operator for the specific Deployment.
|
|
With this feature, you have access to a high level of customizations. However, no guarantee exists that the Deployment will work as expected.
|
|
|
|
The following example illustrates injecting labels, annotations, volumes, and volume mounts:
|
|
|
|
[source,yaml]
|
|
----
|
|
apiVersion: k8s.keycloak.org/v2alpha1
|
|
kind: Keycloak
|
|
metadata:
|
|
name: example-kc
|
|
spec:
|
|
...
|
|
unsupported:
|
|
podTemplate:
|
|
metadata:
|
|
labels:
|
|
my-label: "keycloak"
|
|
spec:
|
|
containers:
|
|
- volumeMounts:
|
|
- name: test-volume
|
|
mountPath: /mnt/test
|
|
volumes:
|
|
- name: test-volume
|
|
secret:
|
|
secretName: keycloak-additional-secret
|
|
----
|
|
|
|
=== Disabling required options
|
|
|
|
{project_name} and the {project_name} Operator provide the best production-ready experience with security in mind.
|
|
However, during the development phase, you can disable key security features.
|
|
|
|
Specifically, you can disable the hostname and TLS as shown in the following example:
|
|
|
|
[source,yaml]
|
|
----
|
|
apiVersion: k8s.keycloak.org/v2alpha1
|
|
kind: Keycloak
|
|
metadata:
|
|
name: example-kc
|
|
spec:
|
|
...
|
|
http:
|
|
httpEnabled: true
|
|
hostname:
|
|
strict: false
|
|
----
|
|
|
|
=== Resource requirements
|
|
|
|
The Keycloak CR allows specifying the `resources` options for managing compute resources for the {project_name} container.
|
|
It provides the ability to request and limit resources independently for the main Keycloak deployment via the Keycloak CR, and for the realm import Job via the Realm Import CR.
|
|
|
|
When no values are specified, the default `requests` memory is set to `1700MiB`, and the `limits` memory is set to `2GiB`.
|
|
These values were chosen based on a deeper analysis of {project_name} memory management.
|
|
|
|
If no values are specified in the Realm Import CR, it falls back to the values specified in the Keycloak CR, or to the defaults as defined above.
|
|
|
|
You can specify your custom values based on your requirements as follows:
|
|
|
|
[source,yaml]
|
|
----
|
|
apiVersion: k8s.keycloak.org/v2alpha1
|
|
kind: Keycloak
|
|
metadata:
|
|
name: example-kc
|
|
spec:
|
|
...
|
|
resources:
|
|
requests:
|
|
cpu: 1200m
|
|
memory: 896Mi
|
|
limits:
|
|
cpu: 6
|
|
memory: 3Gi
|
|
----
|
|
|
|
Moreover, the {project_name} container manages the heap size more effectively by providing relative values for the heap size.
|
|
It is achieved by providing certain JVM options.
|
|
|
|
For more details, see <@links.server id="containers" />.
|
|
|
|
=== Scheduling
|
|
|
|
You may control several aspects of the server Pod scheduling via the Keycloak CR. The scheduling stanza exposes optional standard Kubernetes affinity, tolerations, topology spread constraints, and the priority class name to fine tune the scheduling and placement of your server Pods.
|
|
|
|
An example utilizing all scheduling fields:
|
|
|
|
[source,yaml]
|
|
----
|
|
apiVersion: k8s.keycloak.org/v2alpha1
|
|
kind: Keycloak
|
|
metadata:
|
|
name: example-kc
|
|
spec:
|
|
scheduling:
|
|
priorityClassName: custom-high
|
|
affinity:
|
|
podAffinity:
|
|
preferredDuringSchedulingIgnoredDuringExecution:
|
|
- podAffinityTerm:
|
|
labelSelector:
|
|
matchLabels:
|
|
app: keycloak
|
|
app.kubernetes.io/managed-by: keycloak-operator
|
|
app.kubernetes.io/component: server
|
|
topologyKey: topology.kubernetes.io/zone
|
|
weight: 10
|
|
tolerations:
|
|
- key: "some-taint"
|
|
operator: "Exists"
|
|
effect: "NoSchedule"
|
|
topologySpreadConstraints:
|
|
- maxSkew: 1
|
|
topologyKey: kubernetes.io/hostname
|
|
whenUnsatisfiable: DoNotSchedule
|
|
...
|
|
...
|
|
----
|
|
|
|
Please see https://kubernetes.io/docs/concepts/scheduling-eviction[the kubernetes docs] for more on scheduling concepts.
|
|
|
|
If you do not specify a custom affinity, your Pods will have an affinity for the same zone and an anti-affinity for the same node to improve availability. Scheduling to the same zone if possible helps prevent stretch clusters where cross zone cache cluster traffic may have too high of a latency.
|
|
|
|
=== Management Interface
|
|
|
|
To change the port of the management interface, use the first-class citizen field `httpManagement.port` in the Keycloak CR.
|
|
To change the properties of the management interface, you can do it by providing `additionalOptions` field.
|
|
|
|
You can specify the `port` and the `additionalOptions` as follows:
|
|
|
|
[source,yaml]
|
|
----
|
|
apiVersion: k8s.keycloak.org/v2alpha1
|
|
kind: Keycloak
|
|
metadata:
|
|
name: example-kc
|
|
spec:
|
|
httpManagement:
|
|
port: 9001
|
|
additionalOptions:
|
|
- name: http-management-relative-path
|
|
value: /management
|
|
----
|
|
|
|
NOTE: If you are using a custom image, the Operator is *unaware* of any configuration options that might've been specified there.
|
|
For instance, it may cause that the management interface uses the `https` schema, but the Operator accesses it via `http` when the TLS settings is specified in the custom image.
|
|
To ensure proper TLS configuration, use the `tlsSecret` and `truststores` fields in the Keycloak CR so that the Operator can reflect that.
|
|
|
|
For more details, see <@links.server id="management-interface" />.
|
|
|
|
=== Truststores
|
|
|
|
If you need to provide trusted certificates, the Keycloak CR provides a top level feature for configuring the server's truststore as discussed in <@links.server id="keycloak-truststore"/>.
|
|
|
|
Use the truststores stanza of the Keycloak spec to specify Secrets containing PEM encoded files, or PKCS12 files with extension `.p12`, `.pfx`, or `.pkcs12`, for example:
|
|
|
|
[source,yaml]
|
|
----
|
|
apiVersion: k8s.keycloak.org/v2alpha1
|
|
kind: Keycloak
|
|
metadata:
|
|
name: example-kc
|
|
spec:
|
|
...
|
|
truststores:
|
|
my-truststore:
|
|
secret:
|
|
name: my-secret
|
|
----
|
|
|
|
Where the contents of my-secret could be a PEM file, for example:
|
|
|
|
[source,yaml]
|
|
------
|
|
apiVersion: v1
|
|
kind: Secret
|
|
metadata:
|
|
name: my-secret
|
|
stringData:
|
|
cert.pem: |
|
|
-----BEGIN CERTIFICATE-----
|
|
...
|
|
------
|
|
|
|
When running on a Kubernetes or OpenShift environment well-known locations of trusted certificates are included automatically.
|
|
This includes `/var/run/secrets/kubernetes.io/serviceaccount/ca.crt` and the `/var/run/secrets/kubernetes.io/serviceaccount/service-ca.crt` when present.
|
|
|
|
=== Admin Bootstrapping
|
|
|
|
When you create a new instance the Keycloak CR spec.bootstrapAdmin stanza may be used to configure the bootstrap user and/or service account. If you do not specify anything for the spec.bootstrapAdmin, the operator will create a Secret named "metadata.name"-initial-admin with a username temp-admin and a generated password. If you specify a Secret name for bootstrap admin user, then the Secret will need to contain `username` and `password` key value pairs. If you specify a Secret name for bootstrap admin service account, then the Secret will need to contain `client-id` and `client-secret` key value pairs.
|
|
|
|
If a master realm has already been created for you cluster, then the spec.boostrapAdmin is effectively ignored. If you need to create a recovery admin account, then you'll need to run the CLI command against a Pod directly.
|
|
|
|
For more information on how to bootstrap a temporary admin user or service account and recover lost admin access, refer to the <@links.server id="bootstrap-admin-recovery"/> guide.
|
|
|
|
=== Tracing (OpenTelemetry)
|
|
|
|
Tracing allows for detailed monitoring of each request's lifecycle, which helps quickly identify and diagnose issues, leading to more efficient debugging and maintenance.
|
|
|
|
You can change tracing configuration via Keycloak CR fields as follows:
|
|
|
|
[source,yaml]
|
|
----
|
|
apiVersion: k8s.keycloak.org/v2alpha1
|
|
kind: Keycloak
|
|
metadata:
|
|
name: example-kc
|
|
spec:
|
|
tracing:
|
|
enabled: true # default 'false'
|
|
endpoint: http://my-tracing:4317 # default 'http://localhost:4317'
|
|
samplerType: parentbased_traceidratio # default 'traceidratio'
|
|
samplerRatio: 0.01 # default '1'
|
|
resourceAttributes:
|
|
some.attribute: something
|
|
additionalOptions:
|
|
- name: tracing-jdbc-enabled
|
|
value: false # default 'true'
|
|
----
|
|
|
|
These fields should reflect 1:1 association with `tracing-*` options that contain more information.
|
|
|
|
NOTE: The `tracing-jdbc-enabled` is not promoted as a first-class citizen as it might not be well managed in the future, so it needs to be set via the `additionalOptions` field.
|
|
|
|
For more details about tracing, see <@links.observability id="tracing" />.
|
|
|
|
=== Network Policies (Preview)
|
|
|
|
NetworkPolicies allow you to specify rules for traffic flow within your cluster, and also between Pods and the outside world.
|
|
Your cluster must use a network plugin that supports NetworkPolicy enforcement.
|
|
|
|
The operator can automatically create a NetworkPolicy to deny access to the clustering port of your {project_name} Pods.
|
|
The HTTP(S) endpoint is open to traffic from any namespace and the outside world.
|
|
To enable the NetworkPolicy, set `spec.networkPolicy.enabled` in your Keycloak CR, as shown in the example below.
|
|
|
|
.Keycloak CR with Network Policies enabled
|
|
[source,yaml]
|
|
----
|
|
apiVersion: k8s.keycloak.org/v2alpha1
|
|
kind: Keycloak
|
|
metadata:
|
|
name: example-kc
|
|
spec:
|
|
networkPolicy:
|
|
enabled: true
|
|
----
|
|
|
|
The above example allows traffic from all sources.
|
|
The Keycloak CR can be extended to include a list of rules for each of the endpoints exposed by {project_name}.
|
|
These rules specify from where (the source) the traffic is allowed and it's possible to communicate with the {project_name} Pods.
|
|
|
|
.Extended Network Policy configuration
|
|
[source,yaml]
|
|
----
|
|
apiVersion: k8s.keycloak.org/v2alpha1
|
|
kind: Keycloak
|
|
metadata:
|
|
name: example-kc
|
|
spec:
|
|
networkPolicy:
|
|
enabled: true
|
|
http: <list of rules> #<1>
|
|
https: <list of rules> #<2>
|
|
management: <list of rules> #<3>
|
|
----
|
|
<1> It defines the rules for HTTP endpoint (port 8080 by default).
|
|
Due to security reasons, the HTTP endpoint is disabled by default.
|
|
<2> It defines the access rules for HTTPS endpoint (port 8443 by default.
|
|
<3> It defines the access rules for management endpoint (port 9000 by default).
|
|
The management endpoint is used by the Kubernetes Probes and to expose the {project_name} metrics.
|
|
|
|
The rule syntax is the same as the one used by the Kubernetes Network Policy.
|
|
It makes it easy to migrate your existing rules into your Keycloak CP.
|
|
For more information, check the https://kubernetes.io/docs/concepts/services-networking/network-policies/#behavior-of-to-and-from-selectors[rule syntax].
|
|
|
|
==== Example with OpenShift
|
|
|
|
For a concrete example, let's imagine we have a {project_name} deployment running in a OpenShift cluster.
|
|
Users have to access {project_name} to login, so {project_name} must be accessible from the Internet.
|
|
|
|
To make this example more interesting, let's assume the {project_name} is monitored too.
|
|
The monitoring is enabled as described in the OpenShift documentation page:
|
|
https://docs.openshift.com/container-platform/4.12/observability/monitoring/enabling-monitoring-for-user-defined-projects.html[enabling Monitoring for user defined projects].
|
|
|
|
Based on those requirements, the Keycloak CR would be like this (most parts are omitted, like DB connection and security):
|
|
|
|
.Keycloak CR
|
|
[source,yaml]
|
|
----
|
|
apiVersion: k8s.keycloak.org/v2alpha1
|
|
kind: Keycloak
|
|
metadata:
|
|
name: example-kc
|
|
spec:
|
|
ingress:
|
|
enabled: true #<1>
|
|
networkPolicy:
|
|
enabled: true
|
|
https:
|
|
- namespaceSelector:
|
|
matchLabels:
|
|
kubernetes.io/metadata.name: openshift-ingress #<2>
|
|
management:
|
|
- namespaceSelector:
|
|
matchLabels:
|
|
kubernetes.io/metadata.name: openshift-user-workload-monitoring #<3>
|
|
----
|
|
<1> Enables Ingress for outside access.
|
|
<2> The default OpenShift Ingress class pods are running in `openshift-ingress` namespace.
|
|
We allow traffic from these pods to access the {project_name} HTTPS endpoint.
|
|
The traffic from outside the OpenShift cluster goes through these pods.
|
|
<3> Prometheus pods are running in `openshift-user-workload-monitoring`.
|
|
They need to access {project_name} to scrape the available metrics.
|
|
|
|
Check the https://kubernetes.io/docs/concepts/services-networking/network-policies/[Kubernetes Network Policies documentation] for more information about NetworkPolicies.
|
|
|
|
=== Managing Keycloak Operator Updates (Preview)
|
|
|
|
The Keycloak Operator offers updates strategies to control how the Operator handles changes to the Keycloak CR.
|
|
|
|
[CAUTION]
|
|
====
|
|
While on preview stage, the feature `rolling-updates` must be enabled.
|
|
Otherwise, the {project_name} Operator will fail.
|
|
====
|
|
|
|
**Supported Updates Types:**
|
|
Rolling Updates:: Update the StatefulSet in a rolling fashion, minimizing downtime (requires multiple replicas).
|
|
Recreate Updates:: Scale down the StatefulSet before applying updates, causing temporary downtime.
|
|
|
|
==== Configuring the Update Strategy
|
|
|
|
The update strategy is specified within the `spec` section of the Keycloak CR YAML definition.
|
|
|
|
[source,yaml]
|
|
----
|
|
apiVersion: k8s.keycloak.org/v2alpha1
|
|
kind: Keycloak
|
|
metadata:
|
|
name: example-kc
|
|
spec:
|
|
features:
|
|
enabled:
|
|
- rolling-updates # <1>
|
|
update:
|
|
strategy: RecreateOnImageChange|ForceRecreate|Auto # <2>
|
|
----
|
|
<1> Enable preview feature `rolling-updates`.
|
|
<2> Set the desired update strategy here (Recreate in this example).
|
|
|
|
[%autowidth]
|
|
.Possible field values
|
|
|===
|
|
|Value |Downtime? |Description
|
|
|
|
|`RecreateOnImageChange` (default)
|
|
|On image name or tag change
|
|
|Mimics Keycloak 26.1 or older behavior.
|
|
When the image field changes, the StatefulSet is scaled down before applying the new image.
|
|
|
|
|`ForceRecreate`
|
|
|On any configuration or image change
|
|
|The StatefulSet is scaled down before applying the new configuration or image.
|
|
|
|
|`Auto`
|
|
|On incompatible changes
|
|
|The {project_name} Operator detects if a rolling or recreate upgrade is possible.
|
|
|
|
|===
|
|
|
|
</@tmpl.guide>
|