Performing a schema extraction v4
You can perform a schema extraction using either of the following methods. EDB recommends using the EDB DDL extractor to extract your schemas.
For more information, see Known issues, limitations, and notes.
Download the latest EDB DDL Extractor script from the Migration Portal Projects page.
Migration Portal might become unresponsive for very large SQL files, depending on your system and browser resource availability. To resolve this, try extracting fewer schemas at a time to reduce the SQL file size.
You can run the EDB DDL Extractor script in SQL Developer or SQL*Plus. It uses Oracle’s
DBMS_METADATA built-in package to extract DDLs for different objects under schemas (specified while running the script). The EDB DDL Extractor creates the DDL file uploaded to the portal and analyzed for EDB Postgres Advanced Server compatibility.
You must have
SELECT_CATALOG_ROLE roles and
CREATE TABLE privilege.
Connect to SQL*Plus and run this command:
When prompted, if the Oracle user that is running the script has the required privileges, press Enter to continue running the script.
When prompted, provide the schema name. When extracting multiple schemas, use a comma (
‘,’) as a delimiter.
If you want to extract all the user schemas from the current database, don't mention any schema names while extracting. However, we recommend mentioning the schema names that you want to extract.
When prompted, enter the path or directory for the extractor to store the extracted DDL. If you don't specify a path, the extracted DDL is output to the directory where you ran the SQL*Plus command.
When prompted, enter
nodepending on whether you want to extract dependent objects from other schemas.
When prompted, enter
nodepending on whether you want to extract grant statements from other schemas.
Run the extractor script:
Specify three schemas by separating them with commas:
Enter a path for the output file:
On Linux, you might use:
Similarly on Windows:
yesto extract dependent objects in other schemas:
yesto extract grant statements:
After loading the
edb_ddl_extractor.sql script into SQL Developer and connecting to the source Oracle database, run the script. As the script executes, respond to the prompts.
If the user for the database connection has the required privileges as listed in the Script Output tab, select Yes to continue running the script.
Enter a comma-separated list of schemas, and select OK.
Enter the path for the output file, and select OK. The default is the location of the DDL Extractor script.
noto extract dependent objects, and select OK. The default is to not extract dependent objects.
noto extract grant statements, and select OK. The default setting is not to extract grant statements.
Output of the DDL Extractor run appears in the Script Output tab. The name of the output file appears after the
Extraction Completed message in the script output.
You can also enter single schema name in both SQL*Plus and SQL Developer.
The script then iterates through the object types in the source database. Once the task is completed, the
.SQL output is stored at the location you entered (e.g.,
See file encoding for information about the file encoding format expected by Migration Portal.
The lists and tables that follow show supported and unsupported schemas and objects in Migration Portal.
Exclude these Oracle systems schemas while generating the SQL dump file.
EDB Postgres Advanced Server doesn't support schema names starting with
- Indexes (Except LOB indexes and indexes on materialized views)
- Materialized views
- DB links
- Types and type body
- Role and object grants
COMMENTS on columns, tables, and materialized views are also supported.
- LOB indexes and indexes on materialized views
- XML schemas
- RLS policy
- Queues *
* Even though EDB Postgres Advanced Server provides support for Queue tables, Migration Portal does not currently support it. Queue tables in the source DDL are not uploaded as source and target DDL objects.