...

Introduction

Basically, the synchronization is the mechanism used by Apache Syncope to acquire user/group data from external resources.

Synchronization can be "full" (full reconciliation) or "incremental"...

In the former case, each sync task execution will take over just of changes from the previous execution (if exists and connector permits incremental sync).

In the latter case, each sync task execution will take over of the entire set of data managed by the external resource.

From an external resource to Syncope

...

...

All the entity (user/group) data involved by a synchronization are retrieved from an external resource and processed internally by Syncope itself.

A retrieved entity can be:

1. a matching entity, if a corresponding local/internal entity has been found;
2. or an unmatching entity, otherwise.

By default, Syncope will create locally all the unmatching entities (without linking entities and resources) and will update all the matching ones.

By the way, a different behaviour can be configured working with matching/unmatching rules.

Matching

...

and Unmatching rules

...

Unmatching (corresponding user not found on Syncope):

• IGNORE / UNLINK (do not perform any action);
• ASSIGN (create entity linking the resource);
• PROVISION (create entity without linking the resource).

Matching (corresponding users found on Syncope):

• IGNORE (do not perform any action);
• UPDATE (update matching entity);
• DEPROVISION (delete resource entity);
• UNASSIGN (unlink resource and delete resource entity) ;
• UNLINK (just unlink resource without performing any (de-)provisioning operation);
• LINK (just link resource without performing any (de-)provisioning operation).