Siebel Marketing Guide > Marketing Reference > Understanding Snapshot Generation Internal Processes >

Building a Snapshot File—Rextract and Rcompute


After the first snapshot generation internal process (rcontrol) is complete, the Marketing Server queries the database for columns needed for the snapshot, reads the data from the database, computes custom measures, if required, and writes the results to the snapshot files. In the Marketing Server log file, the information about this step can be found in the section between Application Name - Compute Snapshot and Application Name - Counter Allocate.

This step caches required tables and sorts the merge cursors and accounts for most of the time used by snapshot generation. For information about join methods, see Understanding Joins

The Marketing Server starts two internal processes in this step.

Rextract Process

The steps in this process repeat until you receive the End of Data message.

The following are the steps in the rextract process:

Rcompute Process

The following steps in this process repeat for each customer until you receive the End of Data message.

After the Marketing Server completes caching all cache tables and starts receiving data from the database for merge cursors, the Marketing Server writes the message Joining Records to the Marketing Server log file. The number of denormalized records that were processed is written to the log file by the Marketing Server.


 Siebel Marketing Guide 
 Published: 23 June 2003