Some major limitations due to which partial data loss or structural disparity can happen are listed below.
- The SQL Server endpoint does not support the use of sparse tables.
- Temporal tables are not supported.
- Partition Switching is not supported.
- Column-level encryption is not supported.
- Column descriptions (remarks) are not migrated.
- Replicating data from indexed views isn't supported.
- Collection objects (ex: xml collections) aren’t migrated.
- Fields with the following data types aren’t supported:
- CURSOR
- SQL_VARIANT
- TABLE (user-defined)
- Fields with the following data types are deprecated and are partially supported for backward compatibility reasons. (Suggested to avoid using these data types)
- NTEXT
- IMAGE
- TEXT
- For tables having the below mentioned data types and no primary key, only staging migration can be done, cdc (live-migration) is not possible.
- NTEXT
- IMAGE
- TEXT
- XML
- All MAX data types (like varchar(max), varbinary (max) etc.)
- Presence of schema with the name “cdc” in the source DB will be ignored and not migrated.
- Identities on NUMERIC(p, 0) columns are only supported for p <= 18.
- If p > 18, the pre-migration validation would fail, and the migration task cannot move further.
- For p <= 18, they are converted to BigInt data types.
- Fields with the following data-types are migrated with different data-type on target:
Source Field Data Type
Target Field Data Type
DATETIMEOFFSET
WSTRING