Back up Content Engine
Backing up Content Engine is part of an enterprise-wide backup strategy for IBM FileNet P8, and must be done at the same time as the other P8 components. See Enterprise-wide backup for a list of all components which must be backed up simultaneously.
A Content Engine backup consists of a collection of backups of the Global Configuration Data (GCD) database and one or more object stores. The backups consist of configuration and object metadata as well as content files. The latter can be stored in a database, or on a file system. The backups can also include index storage areas (Verity Collections). File and index storage areas must be kept in sync with any and all associated object stores.
See the following sections to determine the actual files and databases that must be backed up.
The GCD is updated periodically when you modify IBM FileNet P8 domain properties, and when any of the following objects are created or deleted:
You cannot recreate the GCD manually. However, when restoring a system, it is preferable to start from the current GCD if it is available and reconfigure the necessary objects, rather than start from a backup copy.
The GCD is stored in a database. To determine the database used by the GCD, use your application server console to view the data source information for the GCD JNDI DataSource you entered during the Content Engine configuration. (For example, in WebLogic, you will find this listed in the Java™ Database Connectivity (JDBC) data sources. The database information is indicated in the connection pool's URL.) If you do not back up the entire database (for example, if it is shared by other applications), you must back up the FNGCD and FNGCD_AddOn tables.
The object store contains metadata about content, such as the revision number and the physical location of the content in the file system. (The actual content can be stored in the object store database or in a file storage area.) The object store database is updated frequently, whenever object metadata is modified. Some of the actions that cause updates are:
Each object store is stored in a separate database. To determine the database used by each object store, do the following:
Each file storage area is associated with a Content Engine object store. The file storage areas change frequently, whenever changes are made to document contents or annotations.
To determine the location of each file storage area
A new index entry is added for each new document created whose document class is configured for full text indexing.
You can rebuild index collections at any time. However, if the file storage area is large, rebuilding the index collection can take a long time. If you choose to back up the index collections, you must back them up at the same time as you back up the file store in order to keep them synchronized.
Before backing up an index storage area, you must stop all index services on the Content Search Engine that write to the area being backed up. (Refer to the documentation for the Autonomy K2 Dashboard for instructions.) Content Engine logs errors periodically while the K2 index servers are stopped.
NOTE If you have customized the style files on the Content Search Engine, make sure to make a backup copy. See Specify CBR language or the Autonomy K2 Collection Reference Guide (collref.pdf
) for more information.
To determine the location of each index storage area