What is a fallback in Teradata and how is it different from a non-fallback table?

In Teradata, a fallback table is one that has two identical copies of the data stored on separate AMPs (Access Module Processors). This redundancy provides a safety net in case one AMP fails, ensuring data availability and system resilience.

On the other hand, a non-fallback table only has one copy of the data, so it is more vulnerable to data loss in the event of an AMP failure. Non-fallback tables are typically used for non-critical or easily reproducible data.

By using fallback tables, organizations can ensure high availability and data protection in their Teradata environments. It is important to weigh the trade-offs between storage costs and data redundancy when deciding whether to use fallback or non-fallback tables.