FAQ v55

Does Migration Toolkit support the migration of packages?

Migration Toolkit supports the migration of packages from an Oracle database into EDB Postgres Advanced Server. See Functionality overview for information about the migration support offered by EDB Postgres Advanced Server.

Is there a way to transfer only the data from the source database?

Yes. Data transfer is supported as part of an online or offline migration.

Does Migration Toolkit support migration of tables that contain data of the CLOB data type?

Migration Toolkit does support migration of tables containing data of the CLOB type.

Does EDB Postgres Advanced Server support the enum data type?

EDB Postgres Advanced Server doesn't currently support the enum data type but will support it in future releases. Until then, you can use a check constraint to restrict the data added to an EDB Postgres Advanced Server database. A check constraint defines a list of valid values that a column can take.

The following code sample includes a simple example of a check constraint that restricts the value of a column to one of three dept types; sales, admin or technical.

CREATE TABLE emp (
  emp_id INT NOT NULL PRIMARY KEY,
  dept VARCHAR(255) NOT NULL,

  CHECK (dept IN ('sales', 'admin', 'technical'))
);

If you test the check constraint by entering a valid dept type, the INSERT statement works without error:

test=# INSERT INTO emp VALUES (7324, 'sales');

INSERT 0 1

If you try to insert a value not included in the constraint (support), EDB Postgres Advanced Server returns an error:

test=# INSERT INTO emp VALUES (7325, 'support');

ERROR: new row for relation "emp" violates check constraint "emp_dept_check"

Does EDB Postgres Advanced Server support materialized views?

Postgres doesn't support materialized views compatible with Oracle databases. To set up a materialized view/summary table in Postgres you must manually create the triggers that maintain the summary table. Automatic query rewrite isn't currently supported. You must make the application aware of the summary table's existence.

When I try to migrate from a MySQL database that includes a TIME data type, I get the following error: Error Loading Data into Table: Bad format for Time. Does Postgres support MySQL ``TIME`` data types?

Postgres doesn't have a problem storing TIME data types as long as the value of the hour component isn't greater than 24.

Unlike Postgres, the MySQL TIME data type allows you to store a value that represents either a TIME or an INTERVAL value. A value stored in a MySQL TIME column that represents an INTERVAL value can potentially be out of the accepted range of a valid Postgres TIMESTAMP value. If, during the migration process, Postgres encounters a value stored in a TIME data column that it perceives as out of range, it returns an error.

What use cases does the connection retry capability support?

Database scope: The connection retry capability allows Migration Toolkit to reconnect to the target database. Retry attempts for issues with the source database are currently not supported.

Migration scope: This capability allows Migration Toolkit to retry migrating data. Retry attempts for issues with the schema migration are currently not supported.

Modality scope: This reconnection capability is available with the data migration mode (-dataOnly). It's also available for the data migration step when you run a migration without specifying either the -dataOnly and -schemaOnly options.

What do I do if there's an error during a data migration that results in the data for one or more of my tables not being fully migrated?

If Migration Toolkit can't reconnect successfully and one or more of the tables weren't fully migrated, restart the entire data migration (with the -dataOnly option) or configure Migration Toolkit to migrate only the tables that weren't fully migrated in the previous run. You can use connRetryCount, connRetryInterval, and abortConnOnFailure to alter the retry configuration options.

If the source database was accepting write transactions while the previous Migration Toolkit data migration was in process, perform the full (all tables) data migration again to ensure that data migrated to the destination database is in a consistent state.

If you can confirm that no write transactions were performed on the source database while the previous migration was being performed, then it's safe to migrate only those tables that weren't fully migrated.

What do I do if there's an error during a schema migration that results in not all of my schema objects being fully migrated?

Unfortunately, reconnection for schema migration errors isn't supported at this time. If Migration Toolkit can't migrate all schemas, restart the entire schema migration again (with the -schemaOnly option), or configure Migration Toolkit to migrate only the schemas that weren't fully migrated in the previous run.