Setting up full database cacheless search

Siemens Digital Industries Software recommends that you process both the bounding boxes and spatial indexes in batches if the database has a large number of datasets. The generation of bounding boxes and spatial indexes in batches helps prevent the utilities from running out of memory or causing other scalability issues. If you have a small database of less than 50,000 UIDs, you may process all the datasets without any batching.

To deploy cacheless search for a full database strategy:

1. Ensure that you have enabled or upgraded cacheless search with Teamcenter Environment Manager (TEM) before you begin the deployment procedure.
2. Ensure that live background updates are turned off while you manually process the product data.
3. Prepare your data for processing, by updating legacy transforms and removing cycles from the database. This step must be completed carefully or the results of searches may be incorrect.
4. Query for a list of dataset UIDs to process for the entire database.
5. Create the bounding boxes and TSO files for the database.
6. After creating the bounding boxes, regenerate the database statistics to improve search performance.
7. Populate the queue with the entries for the entire database.
8. Process the queue you populated in the previous step. Do not run the qsearch_process_queue utility until the create_or_update_bbox_and_tso has finished.
9. After generating the indexes for a product, regenerate the database statistics to improve search performance.
10. Configure live background updates with Dispatcher or cron jobs to process changes to the indexed data.

Note: In a typical system, you may run a cron job every 15 minutes to process updates. The value could be adjusted, depending on how frequently data changes and the extent of those changes.

Always turn off live background updates if you run the qsearch_process_queue utility to manually process the queue.