Ok...I think I have a better idea of what you need based on the comments above.
First off, if all you need it DB2, I'd go download DB2 10.1 Express-C (the link will take you to whatever is the latest at the time. At the time of this post it is 10.1 Fix Pack 2) to get you started. It is the same binaries as Enterprise Server Edition. All you need to do is purchase a license from IBM and apply it using db2licm -a <license file name>
in order to turn your Express-C into Enterprise Server Edition. Now, that aside, since you are installing this on Windows, make sure you download the appropriate bit version (ie, 32 bit vs 64 bit).
Also, and this from my experience on Windows, make sure you install the 64 bit DB2 into Program Files (no matter which drive you install it to) and if you install the 32 bit DB2, install it into Program Files (x86) (on a 64 bit OS). If you don't do this, Windows won't find the DB2 drivers for any ODBC connections you wish to set up. And even with that, I've had hit and miss luck with 32 bit DB2 on 64 bit Windows.... Best to just use 32 bit DB2 on 32 bit Windows and 64 bit DB2 on 64 bit Windows.
Ok...whew! Now that aside, once you install DB2, it will automatically create an instance named DB2 and most likely create a database called SAMPLE (which you can always remove).
I've found it best (and this is just me) to use a system or local account for the admin ID for DB2. Either way, make sure you let DB2 create the DB2ADMNS and DB2USERS groups. Anyone you want to be able to maintain DB2 will need to be in the DB2ADMNS group and obviously if you just want to access DB2 in the DB2USERS group.
DB2 automatically comes with the DB2 Control Center,which is clunky as all get out. But don't let that turn you off. That is just a GUI, the actual database itself is amazingly fast and powerful. Besides as of version 9.5 or 9.7, IBM has deprecated the Control Center. They still have packaged it with DB2, but they really recommend that you download IBM Data Studio, which is a free, Eclipse-based product that you can use to interact with DB2. I recently downloaded it and tried it. I love it waaaaay better than Control Center. We (the DBAs at our company) are going to work on a project to roll out Data Studio to replace end user and developer access to DB2 to move away from Control Center.
When you download Data Studio, pick the "full product image" under the "Optional Downloads" section. It should be the one about 1.5GB rather than the one above that is only 112MB.
Anyway, hope this gets you started. Let me know if you have other questions.
If you want access to all data (ie, all tables in all schemas), you would need to grant dataaccess.
db2 grant dataaccess on database to user winuser1
If you only want winuser1 to access just the 100 tables in the schema you are referring to, then unfortunately, there is no easy way, you would need to grant SELECT on each table. That being said, it can be accomplished through scripting.
You could do the following
db2 -tnx "select distinct 'GRANT ALL ON TABLE '||
'\"'||rtrim(tabschema)||'\".\"'||rtrim(tabname)||'\" TO USER winuser1;'
from syscat.tables
where tabschema = 'myschema' " >> grants.sql
db2 -tvf grants.sql
This makes use of querying the system catalogs to dynamically generate a script to permission things. This is a lot of how we permission for users we don't want to give dataaccess to.
Here is a good page of the authorities for DB2.
Best Answer
You cannot change a LOB type to a non LOB type and vice versa (I assume LUW here):
http://www-01.ibm.com/support/knowledgecenter/SSEPGG_9.7.0/com.ibm.db2.luw.sql.ref.doc/doc/r0000888.html
[...] SET DATA TYPE altered-data-type Start of changeSpecifies the new data type of the column. The new data type must be castable from the existing data type of the column (SQLSTATE 42837). A LOB column cannot be altered to a different data type (SQLSTATE 42837). A non-LOB column cannot be altered to a LOB data type (SQLSTATE 42837).End of change.
One way is to unload the data with export:
then drop all foreign key pointing to this table, drop the table, create the new table, load the data back to the table and finally add the foreign keys that where dropped.
An alternative approach is to drop the foreign keys, rename the table, create the new table and use load from cursor. When done drop the renamed table