OceanBase Migration Assessment (OMA) is a database compatibility assessment tool provided by OceanBase Migration Service (OMS).
Compatibility assessment
Currently, OMA can assess the compatibility of migration from the following sources to OceanBase AI Database and generate a compatibility report: Oracle, MySQL, PostgreSQL, TiDB, DB2 LUW, RDS MySQL, PolarDB for MySQL, PolarDB for Oracle, DRDS, MSSQL, openGauss, and OceanBase user tenants. For MySQL instances, the report provides the causes of incompatibility and modification suggestions; for other types of instances, only the causes of incompatibility are provided.
OMA supports database object assessment, SQL or PL statement assessment, and full database assessment:
Database object assessment
Directly connect OMA to the source database, automatically obtain information about objects in the source database, and assess the compatibility of OceanBase AI Database with the source database.
OMA can assess the compatibility of DDL statements in Oracle and MySQL databases. It can also assess DDL statements separated with double dollar signs ($$) in text files.
You can use a lightweight client to connect OMA to an Oracle database and export DDL statements for compatibility assessment.
OMA can directly read the DDL statements exported by using MySQLDump for compatibility assessment.
OMA allows you to use db2look to export DDL statements for compatibility assessment.
Assessment of SQL or PL statements
You can directly connect OMA to an Oracle database, scan the V$SQL view to obtain SQL statements of the corresponding schema, and then assess the compatibility of these SQL statements with OceanBase AI Database.
OMA can assess the compatibility of SQL or PL statements that are separated with semicolons (;) or double dollar signs ($$) in text files.
OMA can directly parse SQL statements in MyBatis and iBatis files to assess the compatibility of these statements.
You can directly connect OMA to an Oracle database, periodically scan the V$SQL view over a long period to obtain all SQL statements in a time range, and then assess these SQL statements.
You can connect OMA to a DB2 LUW database, periodically scan the snapshots for dynamic SQL statements over a long period to obtain all SQL statements in a time range, and then assess these SQL statements.
Full database assessment
Full database assessment integrates the object, SQL, and profiling assessment modes, allowing you to complete all three assessments in one task and view the corresponding results in a single report.
Currently, full database assessment in OMA supports different modes for Oracle, MySQL, TiDB, PostgreSQL, and DB2 LUW databases.
Performance assessment
OMA supports SQL replay and stress testing:
You can connect OMA to the source database, collect query-type SQL statements (SELECT statements), and replay them against OceanBase AI Database to verify the compatibility of the SQL statements and the performance of OceanBase AI Database.
OMA allows you to capture the Oracle workload, assess the compatibility of the parsed SQL statements, or replay the SQL statements against OceanBase AI Database to verify their accuracy and performance.
OMA can parse and replay the general log files of MySQL databases.
During an upgrade of OceanBase AI Database, OMA supports pulling query statements from the earlier version and replaying them in the new version.
Export objects and SQL statements from OceanBase AI Database
OMA allows you to export DDL statements for objects in the corresponding schema from OceanBase AI Database.
OMA allows you to export the details of SQL statements in SQL_AUDIT for a specified period from OceanBase AI Database.