Synchronization task types
Note: This topic is relevant for ALM Synchronizer for Agile Manager only. For details about the NextGen Synchronizer see NextGen Synchronizer, or search using the NextGen Synchronizer filter.
If you are still using ALM Synchronizer, we recommend migrating to the more robust NextGen Synchronizer embedded in Agile Manager's configuration area.
The following table describes details about the types of synchronization tasks available: Incremental, and Full.
For details about how to schedule or manually run specific types of tasks, see Run synchronization tasks.
|Incremental synchronizations*||Full synchronizations|
Finds records in each endpoint that were created or modified since the previous synchronization task.
Synchronizes data between the two endpoints according to the mapping and settings you defined for the link.
Performs a full comparison of the records in each endpoint.
This is useful if you want to handle records that were deleted from one of the endpoints, or if you want to synchronize records that were not synchronized in an incremental synchronization.
|List of relevant records||
Each endpoint supplies a list of records that were created or modified since the last synchronization task was run.
The list is compiled by comparing the synchronization task time stamp (the time the last synchronization task was started) to the last time that each record was modified.
Records that were modified after the last synchronization task are included in the list.
Each endpoint supplies a list of all records in that endpoint, and compares the list to the list of records included in the identity mapping table, which lists the records that are already synchronized.
If a record in the list has already been synchronized (is listed in the identity mapping table), and the version of the record in the source endpoint is different than the version of the record in the destination endpoint, Synchronizer considers the record to have been updated since the last synchronization.
Synchronizer performs a cycle redundancy check to determine if the changes include mapped fields, and therefore require synchronization.
If the changes appear in mapped fields, Synchronizer updates the corresponding record in the destination endpoint, depending on the settings defined for the link.
If a record appears in both an endpoint and the identity mapping table for the link, Synchronizer synchronizes the record as it would in an incremental synchronization, based on the link's time stamp, the record version, and cycle redundancy checks.
Some records may be synchronized in a full synchronization that would not be synchronized in an incremental synchronization. This is because all records are requested from each endpoint, and not just those modified since the previous synchronization task.
Incremental synchronizations do not handle deleted records. To handle deleted records, run a full synchronization.
|Records that appear in the identity mapping table but no longer appear in an endpoint are considered to have been deleted in that endpoint, and are handled according to the settings defined for the link, for deleted records.|
Records that had not been previously synchronized are considered to be newly created records.
If you enabled Synchronizer to create records in the destination endpoint based on new records in the source endpoint, Synchronizer creates a new corresponding record in the destination endpoint, and updates the identity map to pair these records.
New records are handled in the same way in both incremental and full synchronizations.
Note: * Incremental synchronizations can sometimes cause data consistency issues. For example, if Synchronizer tries to update a record but is unable to do so, such as if the record was locked in the destination endpoint, the time stamp is updated for the link, although the records have not been synchronized. (Time stamps are updated only when a synchronization task begins.)
Because the time stamp is updated, future incremental synchronizations will not detect that the records in the filter still need to be synchronized.
To handle any data consistency issues, run a full synchronization.