Carrying Case Owner's Manual
82 Oracle TimesTen In-Memory Database Installation Guide
% /opt/TimesTen/tt70/bin/ttMigrate -rx DSN=Salestt70 
sales.dat
Using the ttBulkCp utility
The ttBulkCp utility copies table data between TimesTen data stores 
and ASCII files. The data files used by ttBulkCp can only contain rows 
from a single table. They also do not store the table’s column or index 
definitions. Therefore, when migrating from one TimesTen data store to 
another with ttBulkCp, you must first create the tables and indexes in 
the new data store manually. Then use ttBulkCp to copy the rows from 
the original data store to the new data store. For a description of the 
ttBulkCp syntax and usage, see "Utilities" in Oracle TimesTen In-
Memory Database API Reference Guide.
Note: The release of ttBulkCp must match the release of the data store 
you are copying from or to. In this example, use ttBulkCp Release 6.0 
to save the tables to disk files and use ttIsql and ttBulkCp Release 7.0 
to copy the disk files into the tables of the new data store.
To import data from a data store created with TimesTen6.0:
1. Find all the tables you want to copy into the new release of TimesTen.
2. Use the TimesTen utility ttBulkCp to copy the data in each table to a 
disk file.
3. Define a data source name for the new data store.
4. Use the CREATE TABLE and CREATE INDEX commands with ttIsql 
to recreate each table and index you are importing.
5. Use the TimesTen utility ttBulkCp to copy the contents of the disk 
file(s) into the table(s) of the new data store. If, for example:
• Release 6.0 is installed in: 
/opt/TimesTen6.0/32 and release 7.0 
is installed in /opt/TimesTen/tt70;
• Your DSN for release 6.0 is called 
source600 and your DSN for 
release 7.0 is 
source_tt70.
• You have a ttIsql script named 
create.sql that creates user tables 
and indexes, or use the ttSchema utility to create the SQL statements 
necessary for object creation; and
• You want to migrate the tables ABLE and BAKER from 
source600 
to 
source_tt70.
To copy the tables to disk files, you would execute the commands:










