dwolla/atlassian-crowd repository overview

⁠Atlassian Crowd Docker Image

Docker image with Atlassian Crowd, running on OpenJDK 8 / Alpine Linux with MySQL drivers.

We use a fork of https://github.com/teamatldocker/crowd⁠ as our base image to add specific configurations - mainly enabling SSL on port 8443 and redirecting / to /crowd/ through tomcat. This docker image builds off that and puts configuration specific data in the correct place to allow for an automated deployment of crowd.

⁠Required Environment Variables

⁠DATABASE_CONFIG_OBJECT

Path to an S3 object (e.g. s3://bucket/object.json) structured as follows:

{
  "crowd": {
    "user": "username",
    "password": "password",
    "host": "mysql.database.hostname",
    "port": "3306",
    "database": "crowd"
  }
}

The credentials will be used by the upstream's base image launch.sh⁠ to create a JNDI resource at jdbc/CrowdDS

⁠CROWD_CONFIG_OBJECT

Path to an S3 object structured as follows:

{
  "crowdServerId": "B9AN-B9AN-B9AN-B9AN",
  "license": "AAABGQ0ODAoPeNpdkF1LwzAUhu/plus-some-more-stuff"
}

Both of these values can be found in Atlassian licensing⁠.

⁠CROWD_SERVER_URL

The URL to your Crowd instance, once it’s up and running.

⁠CATALINA_OPTS

Give your Crowd instance more memory by setting CATALINA_OPTS to e.g. -Xms512m -Xmx512m.

⁠CROWD_TLS_SUBJ

The TLS subject for the self-signed certificate generated by the startup script. For example,

/C=US/ST=Iowa/L=Des Moines/O=Dwolla/CN=localhost/emailAddress=your-email-address@localhost"