For configuration, Marche reads all *.conf files in its configuration directory, which is normally /etc/marche (but can be overridden by invoking the daemon with the -c option). Configuration files are in INI format, and all the found files are merged together to form the Marche configuration.

General configuration of marched

The general configuration is usually placed into a single file named /etc/marche/general.conf.


Default: none

The user as which marched will run. Normally, this can be kept empty, so that the root user is used, which is necessary for most jobs. If no user is given, the current user on startup will be used.


Default: none

The group as which marched will run. If no group is given, the current group on startup will be used.


Default: /var/run

The directory where the daemon’s pid file ( will be stored.


Default: /var/log

The directory where all the log files will be stored, in a subdirectory called marche and split by day.


Default: xmlrpc, udp

A comma-separated list of the interfaces that should be started. A list of all available interfaces is given in Available interfaces.


Default: display

A permission level (see the description of permissions) to assign to users that are not authenticated.

Can be the special none level to disable everything for these users.

Interface configuration

The interface configuration is usually also done in general.conf.


Each section called provides configuration for an interface with the name xxx. (To be used, interfaces must be enabled in the interfaces value in the [general] section.) For details about configuring the different interfaces, see the interface documentation.

Authenticator configuration

The management of user/password pairs and their respective permission levels is done by authenticators, of which multiple types can be configured in sections.

If no authenticators are configured, the daemon will accept any clients and assign them the highest permission level.


Each section called configures an authenticator with the name xxx. For details about different authenticators, see the authenticator documentation.

Job configuration

The configuration for individual jobs is usually placed in separate files.


Each section called configures a job called xxx.

The job name is arbitrary, but many jobs use it as a default for other configuration parameters. For example, for the init job, which controls services via an init script, the job name is the default value for the name of the init script.

Each job section must have a value named type, which selects the type of job to provide. A list of available job types, and their configuration parameters, is given in Available jobs.