Obevo will integrate w/ DBMS-provided tooling for reverse-engineering. We prefer this route over using external tools as the DBMS vendors are more likely to keep up-to-date on their own DDL generation tooling.
Obevo will either A) call the vendor APIs directly or B) provide you the command/s to run to generate some interim artifacts for Obevo to then complete processing.
The following tools are currently supported.
| DBMS | Tooling Leveraged by Obevo | Who will invoke the vendor API? |
|---|---|---|
| Sybase ASE | ddlgen | User |
| DB2 | DB2LOOK | User |
| Postgres | pg_dump | User |
| SQL Server | Microsoft.SqlServer.Management.Smo.Scripter class in Powershell | User |
| Oracle | DBMS_METADATA API via JDBC | Obevo |
The tooling generally works as follows: A) If Obevo invokes the vendor API 1. Execute the Obevo command. This will invoke the vendor API and save its output into an interim file, and then proceed w/ the rest of the reverse engineering 2. If the reverse-engineering looks good, then you are done 3. Otherwise, modify the interim file and rerun the Obevo command w/ the interim file argument passed in
B) If the user invokes the vendor API 1. Execute the Obevo command to generate the DBMS-specific commands to reverse-engineer the DDLs to a particular format 2. Execute those DBMS-specific commands to generate the DDL output file 3. Re-execute the Obevo command w/ the DDL output file as an additional argument
%OBEVO_HOME%\bin\deploy.bat NEWREVENG -dbType DB2 -dbSchema YourSchemaName -mode schema -outputPath h:\reveng-example-output -dbHost yourhost.me.com -dbPort 1234 -dbServer MYDB01 Detailing the arguments: -dbType: required, specify SYBASE_ASE or DB2 -dbSchema: required, the schema/database you are querying from -mode: required, use the "schema" value -outputPath: required, specify where your reverse-engineered output should go Connection arguments - specify either -dbHost and -dbPort for Sybase or -dbHost and -dbPort and -dbServer for DB2 Optional Arguments: -debugLogEnabled: use this if you have questions on reverse-engineering that you want to report to the Obevo team. This adds debug information that is useful for support
This will generate the reverse-engineered output under <outputDir>/final
Once you have these files, do the final touches on them as you see fit (e.g. delete junk tables)
Some files that you should note / cleanup
Note that we explicitly don’t include the grants here. This is because you can (and should) use the global permissioning functionality instead.
Once done, return to the Existing Onboarding Guide to continue the onboarding process.
Check with your database administrators on how to obtain the binaries to run ddlgen
pg_dump is available within the core distribution of PostgreSQL. You can use that distro as a client to connect to your DB; you do not need PostgreSQL installed on your computer.