Groups provide structure to the workers within an instance of a process client.
The Folder node type presents group information in four sections.
Group structure
Within the process client, all workers are within a single workers folder. Each worker can link to one or more groups which are in a groups folder. Groups can be associated to a connection (through the connection's Connection folder link), or not attached to a connection. Groups attached to a connection are called "connection groups". Groups are arranged in a single hierarchy (each group links to a higher group through its Parent link, which provides a hierarchy through which workers can be accessed. Each group has a Path reference, and the Path reference fields of the group and its parents are used to derive a unique Folder path for the group. Path reference must be unique amongst groups with the same parent. Groups are by default displayed in name sequence; use Sequence to override this and alter the display position of a group within its parent.
Because workers can be in more than one group, workers can appear in more than one part of the hierarchy. In a typical scenario, a partner might be a supplier and the work of the organisation may be split into projects. An individual worker is visible within a project hierarchy of groups, but is also visible through the group that is associated with the supplier connection. This provides two ways to view workers: through a project hierarchy, or through the group that represents the supplier.
Roles and connections
As well as possibly being a connection group (because it is linked to by a connection's Connection folder field), a group can use the Roles and connections field identify connections and give them roles. These are used to lookup connections by role when creating new workers and when resolving connections in calls. For example, an "auditor" role could be associated with a connection to an audit organisation. Any requirement for an "auditor" for workers in that group, or in sub groups, could use that connection when they need an auditor.
If the default (null) link is used in Roles and connections, which is named "Group role", this establishes a role for the group itself. This can be used for connection groups, or parent groups of these, to indicate the role that the connections take, such as "supplier".
Some roles are used by standard components:
| process or process.role |
Connections that can be queried for definitions of new processes, effectively the process providers for the group. If a role suffix is added, this is passed on to the process provider to filter the processes that should be offered when quering definitions. |
| subscribe or subscribe.entity |
Connections to which published data should be passed. See Publish. If an entity suffix is added, only data for that entity is passed. |
If the group is a connection group, you can use the Address book field to identify groups within the partner's group hierarchy to which processes should be sent. Each group has a name and a path. Some features, like the start new process, will then offer you this list of paths for roles that are suffixed with a *. (Within the process flow, these paths are typically in the initialisation data, and can be accessed using %{data:init.connectionPaths.role}, where role is the role for which the path was entered.)
Appearance
The Icon and Colour for the group can be set.
If the Show Action/Recent/All indicator is set, the list of workers is shown using three tabs which allow the list of workers to be filtered to show only active workers; only active workers plus ones with recent activity; or all workers.
If the Show processes from sub folders indicator is set, add to the list of workers any workers that are active or have recent activity from sub groups.
Permissions
Set Allow task creation to allow new workers to be created and linked to this group.
Set Private to prevent new workers created in this group from being visible to the worker's connection's connection groups. Set this to reduce visibility of the workers created in the group, preventing them from also appearing when looking at the groups that are associated with the worker's connections.
In User groups, select the groups of users that may use this group and the workers within it. User groups are inherited down the group hierarchy. By design, permissions can only be set at the group level, not the individual worker level, to prevent detailed permissions that are difficult to control.
Set Trusted in connection groups to indicate that the associated connection is trusted. This means, amongst other things, that the connection can send processes with steps other than those listed as safe. This would typically be set for connections to Process modules, and not otherwise set.