Installation and Configuration of Keycloak Server The Keycloak Server has two downloadable distributions. keycloak-appliance-dist-all-&project.version;.zip keycloak-war-dist-all-&project.version;.zip
Appliance Install The keycloak-appliance-dist-all-&project.version;.zip is quite large, but contains a complete server (backed by Wildfly) that runs out of the box. The only thing you'll have to enable and configure is SSL. Unzipping it, the directory layout looks something like this: keycloak-appliance-dist-all-&project.version;/ keycloak/ bin/ standalone.sh standalone.bat standalone/configuration/ keycloak-server.json themes/ examples/ docs/ The standalone.sh or standalone.bat script is used to start the server. After executing that, log into the admin console at http://localhost:8080/auth/admin/index.html. Username: admin Password: admin. Keycloak will then prompt you to enter in a new password.
WAR Distribution Installation The keycloak-war-dist-all-&project.version;.zip contains just the bits you need to install keycloak on your favorite web container. We currently only support installing it on top of an existing Wildfly 8 or JBoss EAP 6.x distribution. We may in the future provide directions on how to install it on another web container like Tomcat or Jetty. If anybody in the community is interested in pulling this together, please contact us. Its mostly Maven pom work. The directory structure of this distro looks like this: keycloak-war-dist-all-&project.version;/ deployments/ auth-server.war/ keycloak-ds.xml configuration/ keycloak-server.json themes/ examples/ docs/ After unzipping this file, copy everything in deployments directory into the standalone/deployments of your JBoss or Wildfly distro. Also, copy everything in configuration directory into the standalone/configuration directory. $ cd keycloak-war-dist-all-&project.version; $ cp -r deployments $JBOSS_HOME/standalone/deployments $ cp -r configuration $JBOSS_HOME/standalone/configuration After these steps you MUST then download and install the client adapter as this may contain modules the server needs (like Bouncycastle). You will also need to install the adapter to run the examples on the same server. After booting up the JBoss or Wildfly distro, you can then make sure it is installed properly by logging into the admin console at http://localhost:8080/auth/admin/index.html. Username: admin Password: admin. Keycloak will then prompt you to enter in a new password. You can no longer run Keycloak on JBoss AS 7.1.1. You must run on EAP 6.x or Wildfly.
Configuring the Server Although the Keycloak Server is designed to run out of the box, there's some things you'll need to configure before you go into production. Specifically: Configuring Keycloak to use a production database. Setting up SSL/HTTPS Enforcing HTTPS connections
Relational Database Configuration By default, Keycloak uses a relational database to store Keycloak data. This datasource is the standalone/deployments/keycloak-ds.xml file of your Keycloak Server installation if you used or in standalone/configuration/standalone.xml if you used . File keycloak-ds.xml is used in WAR distribution, so that you have datasource available out of the box and you don't need to edit standalone.xml file. However a good thing is to always delete the file keycloak-ds.xml and move its configuration text into the centrally managed standalone.xml file. This will allow you to manage the database connection pool from the Wildfly/JBoss administration console. Here's what standalone/configuration/standalone.xml should look like after you've done this: jdbc:h2:mem:test;DB_CLOSE_DELAY=-1;DB_CLOSE_ON_EXIT=FALSE h2 sa sa jdbc:h2:${jboss.server.data.dir}/keycloak;AUTO_SERVER=TRUE h2 sa sa org.h2.jdbcx.JdbcDataSource ]]> Besides moving the database config into the central standalone.xml configuration file you might want to use a better relational database for Keycloak like PostgreSQL or MySQL. You might also want to tweak the configuration settings of the datasource. Please see the Wildfly, JBoss AS7, or JBoss EAP 6.x documentation on how to do this. Keycloak also runs on a Hibernate/JPA backend which is configured in the standalone/configuration/keycloak-server.json. By default the setting is like this: Possible configuration options are: dataSource JNDI name of the dataSource jta boolean property to specify if datasource is JTA capable driverDialect Value of Hibernate dialect. In most cases you don't need to specify this property as dialect will be autodetected by Hibernate. databaseSchema Value of database schema (Hibernate property "hibernate.hbm2ddl.auto" ). showSql Specify whether Hibernate should show all SQL commands in the console (false by default) formatSql Specify whether Hibernate should format SQL commands (true by default) unitName Allow you to specify name of persistence unit if you want to provide your own persistence.xml file for JPA configuration. If this option is used, then all other configuration options are ignored as you are expected to configure all JPA/DB properties in your own persistence.xml file. Hence you can remove properties "dataSource" and "databaseSchema" in this case. For more info about Hibernate properties, see Hibernate and JPA documentation .
Tested databases Here is list of RDBMS databases and corresponding JDBC drivers, which were tested with Keycloak. Note that Hibernate dialect is usually set automatically according to your database, but in some cases, you must manually set the proper dialect, as the default dialect may not work correctly. You can setup dialect by adding property driverDialect to the keycloak-server.json into connectionsJpa section (see above). Tested databases Database JDBC driver Hibernate Dialect H2 1.3.161 H2 1.3.161 auto MySQL 5.5 MySQL Connector/J 5.1.25 auto PostgreSQL 9.2 JDBC4 Postgresql Driver, Version 9.3-1100 auto Oracle 11g R1 Oracle JDBC Driver v11.1.0.7 auto Microsoft SQL Server 2012 Microsoft SQL Server JDBC Driver 4.0.2206.100 org.hibernate.dialect.SQLServer2008Dialect Sybase ASE 15.7 JDBC(TM)/7.07 ESD #5 (Build 26792)/P/EBF20686 auto
MongoDB based model Keycloak provides MongoDB based model implementation, which means that your identity data will be saved in MongoDB instead of traditional RDBMS. To configure Keycloak to use Mongo open standalone/configuration/keycloak-server.json in your favourite editor, then change: to: And at the end of the file add the snippet like this where you can configure details about your Mongo database: All configuration options are optional. Default values for host and port are localhost and 27017. Default name of database is keycloak . You can also specify properties user and password if you want authenticate against your MongoDB. If user and password are not specified, Keycloak will connect unauthenticated to your MongoDB. Finally there is set of optional configuration options, which can be used to specify connection-pooling capabilities of Mongo client. Supported int options are: connectionsPerHost, threadsAllowedToBlockForConnectionMultiplier, maxWaitTime, connectTimeout socketTimeout. Supported boolean options are: socketKeepAlive, autoConnectRetry. Supported long option is maxAutoConnectRetryTime. See Mongo documentation for details about those options and their default values.
EAP6.x Logging Accessing the admin console will get these annoying log messages: WARN [org.jboss.resteasy.core.ResourceLocator] (http-/127.0.0.1:8080-3) Field providers of subresource xxx will not be injected according to spec These can be ignored by editing standalone.xml of your jboss installation: ]]>
SSL/HTTPS Requirement/Modes Keycloak is not set up by default to handle SSL/HTTPS in either the war distribution or appliance. It is highly recommended that you either enable SSL on the Keycloak server itself or on a reverse proxy in front of the Keycloak server. Keycloak can run out of the box without SSL so long as you stick to private IP addresses like localhost, 127.0.0.1, 10.0.x.x, 192.168.x.x, and 172..16.x.x. If you try to access Keycloak from a non-IP adress you will get an error. Keycloak has 3 SSL/HTTPS modes which you can set up in the admin console under the Settings->Login page and the Require SSL select box. Each adapter config should mirror this server-side setting. See adapter config section for more details. external Keycloak can run out of the box without SSL so long as you stick to private IP addresses like localhost, 127.0.0.1, 10.0.x.x, 192.168.x.x, and 172..16.x.x. If you try to access Keycloak from a non-IP adress you will get an error. none Keycloak does not require SSL. all Keycloak requires SSL for all IP addresses.
SSL/HTTPS Setup First enable SSL on Keycloak or on a reverse proxy in front of Keycloak. Then configure the Keycloak Server to enforce HTTPS connections.
Enable SSL on Keycloak The following things need to be done Generate a self signed or third-party signed certificate and import it into a Java keystore using keytool. Enable JBoss or Wildfly to use this certificate and turn on SSL/HTTPS.
Creating the Certificate and Java Keystore In order to allow HTTPS connections, you need to obtain a self signed or third-party signed certificate and import it into a Java keystore before you can enable HTTPS in the web container you are deploying the Keycloak Server to.
Self Signed Certificate In development, you will probably not have a third party signed certificate available to test a Keycloak deployment so you'll need to generate a self-signed on. Generate one is very easy to do with the keytool utility that comes with the Java jdk. $ keytool -genkey -alias localhost -keyalg RSA -keystore keycloak.jks -validity 10950 Enter keystore password: secret Re-enter new password: secret What is your first and last name? [Unknown]: localhost What is the name of your organizational unit? [Unknown]: Keycloak What is the name of your organization? [Unknown]: Red Hat What is the name of your City or Locality? [Unknown]: Westford What is the name of your State or Province? [Unknown]: MA What is the two-letter country code for this unit? [Unknown]: US Is CN=localhost, OU=Keycloak, O=Test, L=Westford, ST=MA, C=US correct? [no]: yes You should answer What is your first and last name ? question with the DNS name of the machine you're installing the server on. For testing purposes, localhost should be used. After executing this command, the keycloak.jks file will be generated in the same directory as you executed the keytool command in. If you want a third-party signed certificate, but don't have one, you can obtain one for free at cacert.org. You'll have to do a little set up first before doing this though. The first thing to do is generate a Certificate Request: $ keytool -certreq -alias yourdomain -keystore keycloak.jks > keycloak.careq Where yourdomain is a DNS name for which this certificate is generated for. Keytool generates the request: -----BEGIN NEW CERTIFICATE REQUEST----- MIIC2jCCAcICAQAwZTELMAkGA1UEBhMCVVMxCzAJBgNVBAgTAk1BMREwDwYDVQQHEwhXZXN0Zm9y ZDEQMA4GA1UEChMHUmVkIEhhdDEQMA4GA1UECxMHUmVkIEhhdDESMBAGA1UEAxMJbG9jYWxob3N0 MIIBIjANBgkqhkiG9w0BAQEFAAOCAQ8AMIIBCgKCAQEAr7kck2TaavlEOGbcpi9c0rncY4HhdzmY Ax2nZfq1eZEaIPqI5aTxwQZzzLDK9qbeAd8Ji79HzSqnRDxNYaZu7mAYhFKHgixsolE3o5Yfzbw1 29Rvy+eUVe+WZxv5oo9wolVVpdSINIMEL2LaFhtX/c1dqiqYVpfnvFshZQaIg2nL8juzZcBjj4as H98gIS7khql/dkZKsw9NLvyxgJvp7PaXurX29fNf3ihG+oFrL22oFyV54BWWxXCKU/GPn61EGZGw Ft2qSIGLdctpMD1aJR2bcnlhEjZKDksjQZoQ5YMXaAGkcYkG6QkgrocDE2YXDbi7GIdf9MegVJ35 2DQMpwIDAQABoDAwLgYJKoZIhvcNAQkOMSEwHzAdBgNVHQ4EFgQUQwlZJBA+fjiDdiVzaO9vrE/i n2swDQYJKoZIhvcNAQELBQADggEBAC5FRvMkhal3q86tHPBYWBuTtmcSjs4qUm6V6f63frhveWHf PzRrI1xH272XUIeBk0gtzWo0nNZnf0mMCtUBbHhhDcG82xolikfqibZijoQZCiGiedVjHJFtniDQ 9bMDUOXEMQ7gHZg5q6mJfNG9MbMpQaUVEEFvfGEQQxbiFK7hRWU8S23/d80e8nExgQxdJWJ6vd0X MzzFK6j4Dj55bJVuM7GFmfdNC52pNOD5vYe47Aqh8oajHX9XTycVtPXl45rrWAH33ftbrS8SrZ2S vqIFQeuLL3BaHwpl3t7j2lMWcK1p80laAxEASib/fAwrRHpLHBXRcq6uALUOZl4Alt8= -----END NEW CERTIFICATE REQUEST----- Send this ca request to your CA. The CA will issue you a signed certificate and send it to you. Before you import your new cert, you must obtain and import the root certificate of the CA. You can download the cert from CA (ie.: root.crt) and import as follows: $ keytool -import -keystore keycloak.jks -file root.crt -alias root Last step is import your new CA generated certificate to your keystore: $ keytool -import -alias yourdomain -keystore keycloak.jks -file your-certificate.cer
Installing the keystore to WildFly Now that you have a Java keystore with the appropriate certificates, you need to configure your Wildfly installation to use it. First step is to move the keystore file to a directory you can reference in configuration. I like to put it in standalone/configuration. Then you need to edit standalone/configuration/standalone.xml to enable SSL/HTTPS. To the security-realms element add: ]]> Find the element <server name="default-server"> (it's a child element of <subsystem xmlns="urn:jboss:domain:undertow:1.0">) and add: ]]> Check the Wildfly Undertow documentation for more information on fine tuning the socket connections.
Installing the keystore to JBoss EAP6 Now that you have a Java keystore with the appropriate certificates, you need to configure your JBoss EAP6 installation to use it. First step is to move the keystore file to a directory you can reference in configuration. I like to put it in standalone/configuration. Then you need to edit standalone/configuration/standalone.xml to enable SSL/HTTPS. ... ]]> Check the JBoss documentation for more information on fine tuning the socket connections.
Enable SSL on a Reverse Proxy Follow the documentation for your web server to enable SSL and configure reverse proxy for Keycloak. It is important that you make sure the web server sets the X-Forwarded-For and X-Forwarded-Proto headers on the requests made to Keycloak. Next you need to enable proxy-address-forwarding on the Keycloak http connector. Assuming that your reverse proxy doesn't use port 8443 for SSL you also need to configure what port http traffic is redirected to.
WildFly Open standalone/configuration/standalone.xml in your favorite editor. First add proxy-address-forwarding and redirect-socket to the http-listener element: ... ... ]]> Then add a new socket-binding element to the socket-binding-group element: ... ... ]]> Check the WildFly documentation for more information.
EAP Open standalone/configuration/standalone.xml in your favorite editor. You need to add redirect-port to http connector element and add the RemoteIpValve valve: ]]>
Configuring Servers from the Subsystem If you are using WildFly or EAP,he Keycloak server is deployed and configured from the Keycloak subsystem. This makes provisioning simpler in a domain environment. It also allows you to create more than one Keycloak server instance inside a single WildFly instance. And, you can upload providers, themes, and server configurations without disturbing Keycloak's auth-server.war.
Manually Creating A Server A Keycloak server can be declared by editing standalone.xml or domain.xml. true auth false auth2 ]]> If you create more than one Keycloak server, you will need to use CLI to fully configure each instance. At the least, you will need to run the update-server-config operation.
Using CLI and CLI GUI with the Keycloak Subsystem Servers can also be added/removed or enabled/disabled at runtime using the CLI or CLI GUI tool. These are tools that ship with WildFly/EAP and also with the Keycloak Appliance installation. See CLI or CLI GUI documentation to learn more about how to start the tools, issue commands, and create CLI scripts. To start CLI with the Keycloak Appliance install: /keycloak/bin ./jboss-cli.sh --gui or ./jboss.cli.bat --gui]]> Your server must be running to start in --gui mode.
Basic CLI Commands Command to add a server in CLI: Because "enabled=true", a new Keycloak server will be immediately deployed. By default "enabled" is set to false. Command to remove a server in CLI: The Keycloak server will be immediately deleted and undeployed. Command to enable or disable a server in CLI: The Keycloak server will be immediately deployed or undeployed, but not deleted.
Uploading extra configuration using CLI The Keycloak subsystem allows you to upload keycloak-server.json, provider jars, and theme jars to a Keycloak server instance. The CLI operations for this are "update-server-config" and "add-provider". You may use CLI, CLI GUI, or CLI scripts for these operations. The following examples are shown using CLI GUI for clarity. To use a new keycloak-server.json file for your server, find your server under the Keycloak subsystem. Then right-click the server, select "update-server-config", and upload your file. If you use the update-server-config operation, you should delete or rename <WILDFLY_HOME>/standalone/configuration/keycloak-server.json. Otherwise, all Keycloak server instances will use this file instead of your uploaded file. To upload a new provider jar or theme jar to your server, find your server under the Keycloak subsystem. Then right-click the server, select "add-provider", and upload your file.
Working with overlays When you upload a provider jar, theme jar, or keycloak-server.json file, you are creating an overlay. That is, the file is "overlayed" onto the Keycloak server at deploy time. There are two additional operations that help you manage these overlays. They are "list-overlays" and "remove-overlay". Here are CLI examples of these operations. /subsystem=keycloak/auth-server=my-auth-server/:list-overlays { "outcome" => "success", "result" => [ "/WEB-INF/classes/META-INF/keycloak-server.json", "/WEB-INF/lib/federation-properties-example.jar" ], } /subsystem=keycloak/auth-server=my-auth-server/:remove-overlay(overlay-file-path=/WEB-INF/lib/federation-properties-example.jar,redeploy=true) { "outcome" => "success", } Notice in the "list-overlays" operation, the full path to the server config is /WEB-INF/classes/META-INF/keycloak-server.json. This is always the uploaded path for an "update-server-config" operation. If you remove this overlay, the Keycloak server will revert to its default keycloak-server.json. If you have a keycloak-server.json file in your <WILDFLY_HOME>/standalone/configuration directory, it will always take precedence over both the default and the overlay.
Adding a Keycloak server in Domain Mode In domain mode, you start the server with the "domain" command instead of the "standalone" command. In this case, the Keycloak subsystem is defined in domain/configuration/domain.xml instead of standalone/configuration.standalone.xml. Inside domain.xml, you will see more than one profile. A Keycloak subsystem can be defined in zero or more of those profiles. In the example below, a Keycloak server named "foo" is defined in the "full" profile. The "full" profile is assigned to the "main-server-group". Every WildFly instance that belongs to "main-server-group" will get an identically configured deployment of the "foo" Keycloak server. All operations discussed earlier are valid for a Keycloak server in a domain. You can enable/disable, upload new keyclaok-server.json, and add provider jars. In the following example, any changes that are made to the "foo" server will be automatically propogated to every instance in "main-server-group".