The necessary configuration parameters are queried during the installation of the packages. Most of them have proper default values you can go with. Some of these configuration values can be changed afterwards using the packaging command line tool dpkg-reconfigure <package-name>. Changing the configuration needs a restart of the affected modules, that might include the platform and all extractors.

 

These are the important configuration values regarding the Platform core modules:

  • mico-base package configures the MICO username, password and home directory: These configuration parameters are used in several ways and therefore should not be changed without being cautious.
    1. A system user is created on Debian with the given password and home directory. All extractor processes will run as this user.
    2. The MICO API uses RabbitMQ for communication. During installation of mico-rabbitmq a RabbitMQ user according to these values is created.  These packages use the credentials to connect to RabbitMQ: mico-broker, mico-configuration and mico-persistence.
    3. Marmotta (mico-marmotta) uses the credentials to create a user for it’s backend database on PostgreSQL.
  • mico-marmotta package holds configuration values for the backend database (hostname, database name). As described above the credentials for the databse are the MICO username and password. The hostname plays a special role as it used as the hostname for all core functions and therefore must be resolveable by remote systems that for example run extractors. The hostname is relevant for the common data storage URL, Marmotta URL, RabbitMQ and Broker services. (We agree that the hostname configuration should not be part of this package as it is confusing. This will be fixed in the future.)
  • The mico-registration-service is responsible for managing the registerations of extractors for the platform. It’s package configuration values affect the database for storing the extractor management data only.
  • The mico-workflow-management-service handles the extraction workflows. The necessary data is stored in a database, where the package configuration handels the database settings.
  • The backend storage that is used by the broker, extractors and all other components using the MICO API can be configured in the JAVA Broker context. You can find the context configuration file at /etc/tomcat7/Catalina/localhost/broker.xml.tmpl and adapt the parameter mico.storageBaseUri to your needs. Don’t forget to do a dpkg-reconfigure mico-broker. All components need to be restarted if changed, as this (new) setting is communicated on startup.

More configuration settings can be found in /etc/tomcat7/Catalina/localhost/*.xml.tmpl.

If you have any troubles you might want to check the log files. The logs of the Platform can be found in /var/log/tomcat7/ and for the extractors /var/log/mico-extractor-<extractor-name>/.