Grouper provisioning translations

The comparisons and logic happen on "target objects" which are translated from grouper

Examples

• This will reverse a name, replace the separators (colons) with something else (in this case dot), and maxlength of 64 will just truncate there.  This will generate a generally unique name

Roles

If you need a group membership or privilege on a user in a translation you can use this (4.11.0+, 5.9.0+).  These return a boolean

If you need list of users from a group based on membership or privilege, can get a set of strings based on: subjectId, subjectIdentifier0, subjectIdentifier1, subjectIdentifier2, email

High level translation

Open

1. Data is retrieved from Grouper, and translated to a standard format

   1. Note: see below, some data might be needed from cache (in DB)

   2. If data is not in cache or there are errors (since cache out of date), lookup in Subject API or Target for users or groups

   3. Abbreviated data is retrieved for Groups/Members/Memberships, but the full sync objects are retrieved

2. Data is retrieved from the Target using the DAO and translated to a "standard" format for that target type (e.g. ldap attributes are put in attributes in the group or entity)

3. Configuration for the grouper provisioning objects translated to grouper target objects.  The target object format is specific to the target and documented in the provisioner

   1. If memberships are translated to group attributes or subject attributes, then there is still a reference from the attribute to the membership so things can be tracked

4. Index the target provisioning objects and the grouper target objects so they can be compared

5. The grouper target objects and target provisioning objects are compared by the framework in a generic way

6. If the target and grouper are not in sync, the changes are collected in the form of target provisioning objects

   1. There is a list of individual changes needed in the target (e.g. add this attribute value to this object)

7. The DAO will call CRUD operations (could be batched) to get the target in sync with Grouper

Subject link

"Subject link" is when the subject source needs to be consulted since data to provision is not just the subject id (need attributes cached in grouper or need more up to date data)

Here is the design for subject link from grouper

Open

1. Data is read from Grouper which includes the "sync" tables

2. For entities, this could have some data cached for a subject
   
   

   provisioner.sqlProvTest.syncMemberFromId3AttributeValueFormat = ${subject.attributes['myLdapId']}

3. If the data is not cached (or if there is an error), look this up in the subject API (will batch multiple subjects at once)

4. Then once this data is in the "grouper provisioning objects", continue with the translation to the grouper target objects

The provisioning framework will see if the "memberFromId3" is null when getting data from sync cache, those subjects will be resolved and that data will be populated and used

This will translate that to the entity id

Target user link

"Target user link" is when the target needs to be consulted since data to provision is not just the subject id or subject api data (need cached in grouper or need more up to date data)

Here is the design for target user link from grouper

Open

1. Data is read from Grouper which includes the "sync" tables

2. For entities, this could have some data cached for a subject
   
   

   # main identifier of the user on the target side provisioner.sqlProvTest.syncMemberToId2AttributeValueFormat = ${targetEntity.attributes['dn']} # identifier of the user as referred to by the group provisioner.sqlProvTest.syncMemberToId3AttributeValueFormat = ${targetEntity.attributes['uid']}

3. If the data is not cached (or if there is an error), look this up in the target (will batch multiple entities at once)

4. Then once this data is in the "grouper provisioning objects", continue with the translation to the grouper target objects

The provisioning framework will see if the "memberToId2" or "memberToId3" is null when getting data from sync cache, those entities will be resolved and that data will be populated and used

This will translate that to the entity id

Target group link

"Target group link" is when the target needs to be consulted since data to provision is not just the group uuid, name, id index, etc data (need cached in grouper or need more up to date data)

Here is the design for target group link from grouper

Open

1. Data is read from Grouper which includes the "sync" tables

2. For groups, this could have some data cached from the target
   
   

   # main identifier of the user on the target side provisioner.sqlProvTest.syncGroupToId2AttributeValueFormat = ${targetGroup.attributes['dn']}

3. If the data is not cached (or if there is an error), look this up in the target (will batch multiple entities at once)

4. Then once this data is in the "grouper provisioning objects", continue with the translation to the grouper target objects

The provisioning framework will see if the "groupToId2" or "groupToId3" is null when getting data from sync cache, those entities will be resolved and that data will be populated and used

This will translate that to the group id

Beans

ProvisioningGroup

ProvisioningEntity (e.g. user)

ProvisioningMembership

ProvisioningAttribute

Grouper auto translated from SQL

Grouper native comes from SQL queries.  This is fairly standard and minorly customizable

Open

Grouper translated to target

The Grouper provisioning format is processed and the "target object" format needs to be reached.

First the grouper provisioning groups are cycled through, and for each group, a grouper target group will be created, and a translation will assign fields of the target object from the grouper objects

If a grouperToTargetTranslation "for" is "group" then that translation will be evaluated while looping through groups.

In this example, use the group "name" as the matching ID of the group

If a grouperToTargetTranslation "for" is "entity" then that translation will be evaluated while looping through entities

In this example use the subject id as the matching ID

If a grouperToTargetTranslation "for" is "membership" then that translation will be evaluated while looping through memberships

Target provisioning objects

Depending on the target, fields will be populated and documented for that provisioner.  Note: any state needed to make changes in the target need to be stored in these beans

Target native objects (if exist)

Index objects

Use the targetGroup, targetEntity, and targetMembership variable to refer to the grouperTargetGroups, grouperTargetEntity's, grouperTargetMemberships, targetProvisioningGroups, targetProvisioningEntity's, and targetProvisioningMemberships

If you are supporting renames, then the id of the targets need to something opaque and unchanging (e.g. grouper group idIndex or uuid).

Note: you only need to specify the id's for objects that are in the target (e.g. you might not need all three unless there are groups, entities, and memberships in the target).

If something is specified then the group id and entity id will be used for targetGroups and targetEntities.  For targetMemberships, it will by default use MultiKey of provisioningGroupId and provisioningEntityId.

The ID cant be null unless its an insert and the target expects a null for insert

The type of the id should be a string, numeric, or a multikey of a few strings / numerics.  The ${ } is optional