Insert Production Data
The topics in this section describe the background processes used to populate the production database with the information in the staging database.
A Batch Process Per Table
An insert to production batch process is provided for each table of a maintenance objects that is eligible for conversion.
The batch process is responsible for transferring to production all the rows from the staging table while replacing all references to old keys with their corresponding new keys as follows:
• Keys residing in prime keys, foreign key fields and characteristics are resolved using the corresponding
old key / new key tables.
• Keys residing in generic foreign key reference tables where each row represents a reference to an entity’s maintenance object and prime key values are also resolved using the corresponding
old key / new key tables. Note that all entities referenced this way must either be convertible or view of production. A reference to an entity that is explicitly marked as not convertible is not supported as it cannot be resolved.
• Keys residing in XML storage fields are resolved using the corresponding
XML resolution tables.
Note:
Foreign keys residing in any other form of storage are not resolved. For example, any references stored in raw elements, log records, or any other field that is not used as an XML storage field.
All insertion batch processes are independent and may run concurrently. Also note, all insertion batch processes can be run in many parallel threads as described in the next section (in order to speed execution).
Recommendations To Speed Up Insertion
The following points describe ways to accelerate the execution of the insertion batch processes:
• Before running the first insertion batch process:
◦ Rebuild the index on the prime key on the old key / new key table.
◦ Re-analyze the statistics on the old key / new key table.
◦ Alter all indexes on the production tables being inserted into to be unusable.
• After the insertion programs have populated production data, rebuild the indexes and compute statistics for these tables.
Note:
In a Cloud installation, refer to the “Data Conversion Support for Cloud Implementations” for more information about the various tools provided to support database related conversion tasks.
Parent topic