When documents are checked in and out of the system, a new version is created each time, except where the check-in method is configured to replace the previous revision. This process can lead to the accumulation of many superseded versions that may be of no use to the users. Consider the following options that will reduce the number of files stored by the system:
-
Purge old versions of released revisions. This operation can be done from the document revision or master. It removes information from the database, reducing its size, and removes files from the vault.
-
Configure the document sign off method to remove the superseded versions automatically.
-
Change the check in method to replace the previous version, but only if there is no need to ever revert to the superseded version.
In many cases, reducing the number of document versions stored in the system is not an option. The following items are things to consider to improve the response time when querying for documents without removing documents from the system.
-
Check the column set used in the list view to display the query results. When large amounts of data are returned, this can have a significant impact. Each relationship displayed in a column requires a relationship expansion on the server and the return of the complete object to the client. If there are many relationships in the list view, remove them and see if performance improves. Consider the performance trade off when evaluating relationships. For example, does the convenience of showing the files attached to the documents outweigh the effect on performance?
-
Check the user’s role to see what domains are being searched.
The fewer domains or domain groups you search against, the fewer database tables
are searched and the better the performance. For example, when the system is configured
for authoring, make sure that only the domains containing data relevant to the engineer's
discipline are on the relationships between the role and the access groups.