Siebel EIM is a server component that transfers data between the Siebel database and other data sources. This exchange of information is accomplished through interface tables called EIM tables. The EIM tables act as a staging area between the Siebel application database and other data sources. EIM is the primary method of loading large volume of data into the Siebel database and should be used perform bulk imports, updates, merges, and deletes of data.
The process will be to
Load data into EIM tables.
Run Siebel EIM process to import the data from the EIM tables into the Siebel base tables.
Verify that all indexes exist for the tables involved. However, that for large loads one should drop most of the indexes from the target tables to increase the speed of the process, rebuilding those indexes afterward when the process is finished.
Limit tables and columns to be processed using ONLY BASE TABLES/COLUMNS configuration parameters to minimize EIM processing.
Disabling the Docking: Switching off transaction logging improves performance.
Altering batch sizes to find the optimal batch size for a given business component typically helps resolve performance issues. The batch size is dependent upon the quantity of data and which type of EIM process you are running.
NOTE: Although the limit of rows one can process is directly related to the capabilities of the database server, executing batches greater than 100,000 rows is discouraged.
For EIM delete processes that use the DELETE EXACT parameter, use a batch size of 20,000 rows or less.
Try using batch ranges (BATCH = x-y). This allows the run with smaller batch sizes and avoid the startup overhead on each batch. The maximum number of batches that you can run in an EIM process is 1,000.
Perform regular table maintenance on EIM tables. Frequent insert or delete operations on EIM tables can cause fragmentation.
Delete batches from EIM tables on completion. Leaving old batches in the EIM table wastes space and could adversely affect performance.
Run independent EIM jobs in parallel.
Set the USING SYNONYMS parameter to FALSE in the .IFB file to indicate that account synonyms do not need to be checked.
Use the SQLPROFILE parameter to identify slow-running steps and queries.