Optimizing the NIS database

When implementing the Nautical Solution in an organization, there are several ways to maintain your system so that performance is not degraded over time.

The NIS is the database that most likely will have a large number of table row inserts, modifications, and deletes. Over time, performance degradation will be a concern if not handled properly by the administrator. Since the data content of the NIS varies in scale, the features that exist in the NIS can vary significantly in the geographic size, making this database a perfect candidate for spatial index problems if not handled appropriately.

The following is a list of maintenance tasks the ArcSDE administrator perform on the NIS database to ensure optimal performance. The order of operations for a maintenance plan should be compress, rebuild indexes, analyze. This can be scripted and run via scheduled processes, depending on your operating system.

Schritte:
  1. Delete inactive versions when possible.
    1. Right-click the database connection in the Catalog Tree window and click Versions.

      The Version Manager dialog box appears.

    2. Right-click a version other than the DEFAULT version and click Delete.
    3. Click Yes on the Delete Version message box that appears.
    4. Repeat steps b–c for any other versions that are not the DEFAULT version.
    5. Close the Version Manager dialog box.

    After data loading using the S-57 to Geodatabase importer, the cell's contents exist in an import version (IMP_). All data is held in tables known as delta tables, which are typically slower for the system to query. By deleting these versions after you reconcile and post to DEFAULT, you ensure that these versions will not slow down the system when a compress is performed.

  2. Batch reconcile versions every night, if possible.
  3. Compress your database often, especially after reconciling and posting edits.

    This can be done via the Catalog window when connected as the SDE administrator user, via geoprocessing, or via commands. You can schedule a compress whenever there is downtime, such as overnight or on weekends.

  4. Rebuild attribute indexes.

    You must be connected as the data owner.

  5. After loading a significant number of cells that vary in scale band, recalculate your spatial index.

    Do this again only after a significant number of new features are added to the database. You must be connected as the data owner.

  6. Analyze the tables and feature classes with many records.

    Always analyze the Nautical feature dataset, PLTS_COLLECTIONS, PLTS_FREL, and any other tables with many records. You must be connected as the data owner.


9/17/2010