Convert Oracle-to-SAP HANA with Advanced SQL Migration Tool

1. Download the script provided in the link below (APPDEMO2.sql) and copy it to a local folder (e.g. c:\sql_scripts\APPDEMO2.sql) in a host where you have the Oracle client installed and access to the Oracle instance:

   https://d.dam.sap.com/a/cxJYuQ3

2. Connect to the Oracle database with the sqlplus utility and run the script with @ directive:

   @ c:\sql_scripts\APPDEMO2.sql

   

1. Create what will be the project folder, for example, C:/tmp/Advanced_SQL_Mig_tutorial.

   

2. Open a migration tool command line with C:\cygwin\start_bash.bat.

   Check the installation guide if needed (see Prerequisites).

3. From the command line, execute the migration tool utility sapdbmtk with config option, specifying the full path to the project folder:

   sapdbmtk config C:/tmp/Advanced_SQL_Mig_tutorial

4. The Source/Targe DBs configuration window appears. Select Oracle as source DB and SAP HANA with the corresponding version as target.

   

5. Click Save & Continue.

   A popup window appears asking you to save the selected databases. Click OK and the migration tool configuration window for this particular project appears.

   

6. Set the following options as indicated (if you want to get information for an option click the corresponding question mark):

   • ID # 110 > IP address for the host (or hostname) where the gateway will run (for simplicity the gateway can run on the same host as the migration tool is running)

   • ID # 111 > port where the gateway will listen (e.g., 12345)

   • ID # 112 > set to 600

   • ID # 113 > set to no if you want to keep gateway up when extract command finishes

   • ID # 120 > set to in ('APPDEMO2') (this will make the extract command get only objects under APPDEMO2 schema)

   • ID # 201 > set the default password you want to use for users to be created in SAP HANA

   • ID # 210 > set to yes

   • ID # 212 > increase the default value to 150

   • ID # 220 > set to INT

   • ID # 221 > set to 300

   • ID # 222 > set to numeric (8,2)

   • ID # 223 > set to INT

   • ID # 300 > set to direct_select

   • ID # 330, # 331, # 332 > set to lowercase

   • ID # 383 > set to warning

   • ID # 740 > tenant DB name in SAP HANA

   • ID # 751 > host where SAP HANA is running

   • ID # 752 > port where SAP HANA is running

   • ID # 753 > SAP HANA username, for simplicity use SYSTEM to avoid having missing permissions

   • ID # 754 > SAP HANA user password for user specified in previous parameter

   • ID # 757 > set to no

7. Click Validate & Save, and a popup window appears to confirm the changes.

   Click OK and then Exit to finish the configuration window.

The project folder is configured now, and a subfolder called config is created.

Every migration tool command creates a log file located under subfolder called reports (the folder is created automatically if it doesn’t exit).

Data model and SQL objects extraction can be done either by connecting to the source database via the migration tool gateway or using text files with object definitions (DDLS). Section 1 is mandatory and you will go through the extraction via the migration tool gateway. Then section 2 is optional, do it if you want to see the extraction using a text file as well.

IMPORTANT: It is recommended for you to try both methods, go through the gateway first in section 1 and then go through optional steps in section 2

The migration tool gateway is usually executed in a separate host, but for simplicity in this tutorial it will be executed in the same host where the migration tool is installed, notice that Java version 1.8 minimum is required to run the gateway.

You are going to configure and start the gateway now.

1. Open config.properties file under c:\sapdbmtk\gateway folder and modify lines 4 and 5 as described in the picture below.

   The bindaddress should be the ipadress/hostname for the host where the migration tool is installed, and the port should be the one you have specified in configuration setting with ID #111 (12345 as specified in step number 2).

   

2. Oracle JDBC driver has to be installed in the host where the gateway runs – in this case it is installed under c:\sapdbmtk\Gateway\ora_jdbc (ojdbc7.jar).

   

3. Create a .bat file called START_SAPDBMTK_Gateway_oracle.bat under c:\sapdbmtk\gateway to start the gateway, you can use a variable to set the full path to the JDBC driver, the content of the file should be this:

   Code Copy

   set ORACLE_JDBC=C:\sapdbmtk\gateway\ora_jdbc\ojdbc7.jar
   start_sapdbmtk_gateway %ORACLE_JDBC%

   

4. Start the gateway executing the new START_SAPDBMTK_Gateway_oracle.bat file – a new command line window will be opened. Check that the ipadress/hostname and port are the right ones.

   

   You are ready to run the migration tool extract command.

5. Once the gateway is up and running execute the migration tool utility sapdbmtk with extract option, the command line parameters you have to specify are:

   sapdbmtk extract <project_folder> <ora_intance> <ora_host> <ora_port> <ora_user> <passwd>

   For example:

   c://sapdbmtk 32# sapdbmtk extract C:/tmp/Advanced_SQL_Mig_tutorial <ora_intance> XXX.XXX.XXX.XXX 1521 <oracle_user> <system_passwd>

   NOTE: the Oracle user you specify for this command needs to have read permissions on the catalogs, it will not perform any modification in the source database, SELECT_CATALOG_ROLE is enough for this purpose.

   IMPORTANT: windows firewall or antivirus software can prevent the migration tool connecting to the gateway, if you have connectivity issues try stopping/configuring firewall/antivirus

6. When you execute the extract option for the first time in a new project folder you will get 2 popup windows to confirm that compliance with 3rd party rights has been verified. When each of those 2 popup windows appears just read and click on I confirm, then rerun the command again (you will have to rerun the command twice the first time).

   

7. Execute the extract command, you will see something like this.

   

   The extract command is done when you get the prompt again.

   

   The complete output for the extract is located under <project_folder>\reports in a file named .SQL-Extraction.Oracle-HANA.YYYYMMMDD.hhmiss.log

   With the extract command the migration tools builds its own metadata about the data model and SQL objects found in the source database. A new subfolder called extracted is created under the project folder, in it you will find a file for every SQL object like a stored procedure/function/trigger. and another subfolder called catalog storing files with information about users, tables, columns and so forth.

   

Before proceeding with this optional step you have to delete subfolder called extracted under the project folder, it was generated when running extract option with the Gateway and will generate it again now

1. DDL file/s preparation.

   The script that is provided and used to create the Oracle user and data model in step 1 can be used for this purpose as well.

   Copy that script (APPDEMO2.sql) to the Windows host where the migration tool is running, for example, under folder c:\sql_scripts\APPDEMO2.sql.

2. Migration tool extract_offline command.

   Migration tool extract_offline command is used to perform the extract using a text file, this command has the following command line parameters:

   sapdbmtk extract_offline <project_folder> '<full_path_to_text_file>'

   Execute this command with the right parameters in your case:

   sapdbmtk extract_offline C:/tmp/Advanced_SQL_Mig_tutorial 'c:/sql_scripts/APPDEMO2.sql'

   The output when it starts will be something like this:

   

   Once the command is finished you will see the following:

   

   Subfolder extracted is created under the project folder in the same way it was created for the extract command in section 1.

Once the source data model and SQL object are extracted, the next step is to run the conversion to the target database – SAP HANA in this case. For that, you have to use the migration tool convert command with the following parameters:

sapdbmtk convert <project_folder>

For example:

c://sapdbmtk 32# sapdbmtk convert C:/tmp/Advanced_SQL_Mig_tutorial

When you run the convert command, you can see this output in the command line window.

And once it is completed, you will get the prompt with this output:

The complete output for the extract is located under <project_folder>\reports in a file named SQL-Conversion.Oracle-HANA.YYYYMMMDD.hhmiss.log.

When convert is executed, 2 new subfolders are created in the project folder: converted_orig and scripts.

• In converted_orig you can find a .sqlscript file with the converted code for each procedure/function/trigger and a .convertlog file per object as well with conversion details.

• In scripts you can find all the ready-to-run shell scripts and SQL scripts generated by the migration tool to be directly executed against the target database. Notice that a copy of converted_orig folder is created under scripts with name converted.

After running the convert command two shell scripts called run_ddl_scripts.sh and run_sql_scripts.sh. have been created under <project_folder>\scripts, they are ready to be executed against SAP HANA. The former will create the data model (tables/views/indexes/synonyms …etc.) and the latter will create SQL objects like stored procedures, functions, views,…etc.

Below you see how they look like, this screen shot belongs to run_ddl_scripts.sh but run_sql_scripts.sh is similar. Notice that the connection parameters specified in configuration settings #740, #750, #751, #752, #753 and #754 (this was done in step number 2) are used in these 2 shell scripts.

1. Execute run_ddl_scripts.sh, which will give you something like this:

   

   When it is complete, you will get the prompt and an output similar to this:

   

   Note: If you are using SAP HANA Cloud you will get some warnings when executing this script.

2. Execute run_sql_scripts.sh, which will give you something like this:

   

   When it is complete, you will get the prompt and an output similar to this:

   

   Note: You will get some warnings as well if you are using SAP HANA Cloud in the same way you got them when running run_ddl_scripts.sh.

   After running these two shell scripts, the data model and SQL objects have been created into SAP HANA, you can check it with SAP HANA Studio.

   

Run some SQL stuff in Oracle to force a couple of errors and see how Oracle behaves, then run exactly the same in SAP HANA to confirm it is working similarly and showing the same behavior after the conversion.

1. Oracle

   Open the provided text file with the Oracle data model (APPDEMO2.sql) and check the SQL code for a stored procedure called appdemo2.insert_author.

   Notice that it is checking different things before performing the real insert into appdemo2.authors table and raising errors when apply – for example, it checks that state value is no longer than 2 characters and phone has an specific format with no letters (it is being validated using an Oracle function).

   

   Connect to your Oracle instance with the sqlplus utility and run the following procedure call:

   SQL Copy

   call appdemo2.insert_author (
   'Javier',
   'Martin',
   '+34 a666 666',
   'C/ Torrelaguna Nº 77',
   'Madrid',
   'MAD',
   'Spain',
   '28043',
   null
   )
   /

   It fails with the following error:

   

   Modify the stored procedure call by removing last letter (D) from state field:

   SQL Copy

   call appdemo2.insert_author (
   'Javier',
   'Martin',
   '+34 a666 666',
   'C/ Torrelaguna Nº 77',
   'Madrid',
   'MA',
   'Spain',
   '28043',
   null
   )
   /

   Run the new stored procedure call, and a second error appears:

   

   Modify the stored procedure call again by removing character a from phone field:

   SQL Copy

   call appdemo2.insert_author (
   'Javier',
   'Martin',
   '+34 666 666',
   'C/ Torrelaguna Nº 77',
   'Madrid',
   'MA',
   'Spain',
   '28043',
   null
   )
   /

   Run the stored procedure call again. Now it works well and the author is created.

   

2. SAP HANA

   Now you are going to repeat the same procedure calls against SAP HANA and see that behavior is similar to Oracle.

   Connect to your SAP HANA instance with hdbsql using the following parameters:

   hdbsql -u SYSTEM -p XXXXXX -n <ipdress>:<port> -d <SID> -j -A -m -fn -a -cgo -strictseparatorline

   Run the following stored procedure call:

   SQL Copy

   call appdemo2.insert_author (
   'Javier',
   'Martin',
   '+34 a666 666',
   'C/ Torrelaguna Nº 77',
   'Madrid',
   'MAD',
   'Spain',
   '28043',
   null
   )
   go

   Notice that the same error you had in Oracle is appearing in SAP HANA as well.

   

   Modify the stored procedure call by removing D from state field:

   SQL Copy

   call appdemo2.insert_author (
   'Javier',
   'Martin',
   '+34 a666 666',
   'C/ Torrelaguna Nº 77',
   'Madrid',
   'MA',
   'Spain',
   '28043',
   null
   )
   go

   Run the stored procedure call again and the second error pointing to an invalid phone appears.

   

   Modify the stored procedure call by removing character a from the phone number:

   SQL Copy

   call appdemo2.insert_author (
   'Javier',
   'Martin',
   '+34 666 666',
   'C/ Torrelaguna Nº 77',
   'Madrid',
   'MA',
   'Spain',
   '28043',
   null
   )
   go

   Run the stored procedure again, you see now that it work properly and the new author is created