Du hast den Nagel auf den Kopf getroffen.
Hierzu gibts bei publib.boulder.ibm.com auch ein PDF-Handbuch.

Auszug:

A simple example of making text searchable
1. At your command entry display, enter:
CALL(QDB2TX/DB2TX)
2. Enable the server for text search
To enable the connected server, enter:
db2tx=>ENABLE SERVER FOR DB2TEXT
3. Enable a text table for text search (optional)
You need to enable a text table only if you want to create a single index for the
whole table.
4. Enable a text column for text search
Enter the following command to enable DB2 Text Extender to search in text
column mycolumn in table db2tx.sample, and to assign the name myhandle to the
handle column that this command creates.
db2tx=>ENABLE TEXT COLUMN db2tx.sample mycolumn HANDLE myhandle
This command creates a text index. Default values are used for the type of
documents being indexed and for the index characteristics.
5. Check the status of the index you are creating
Enter:
db2tx=>GET INDEX STATUS db2tx.sample HANDLE myhandle
6. Leave the DB2 Text Extender command line processor
Enter:
db2tx=>QUIT
7. Start the interactive SQL session
At the command entry display, enter:
STRSQL NAMING(*SQL)
8. Search for text
Now your documents can be searched. Try this SELECT command which finds
all occurrences of searchterm in the text that you have just indexed:
SELECT COUNT (*)
FROMsample
WHERE DB2TX.CONTAINS (myhandle,'"searchterm"') = 1