-
Query sets may be applied to an existing Alfresco repository. For example a query set could be applied to a system which has been upgraded to 7.0.0 that already contains hundreds of millions of documents.
-
Multiple (zero or more but typically not more than 10) query sets may be defined. Each will have its own name. It will be possible to replace a query set with a new version or to remove it completely. The definition can include the properties or aspects applied to nodes and if necessary (for selected databases) the order of columns in compound indexes. Query sets are defined using JSON files.
-
Administrators perform a query set refresh in the Admin Console. The addition of new query sets, the replacement of an existing query set or complete removal does not require a restart, an outage or have a major impact on normal operations. The alfresco.log will contain messages to reflect progress. When a new query set is identified, the system will start populating a denormalized table in the background. It will also abandon the table population before it is complete, if a new version of the query set is created. The implementation will also need to identify a query set or if a previous version is no longer needed and issue a message to the alfresco.log to advise that the query set can be deleted.
-
Once a denormalized table has been created and fully populated, it will automatically start being used.
-
The Query Accelerator will provide ATOMIC (transactionally consistent) results.
-
The Query Accelerator is only an enterprise edition feature.