You can repartition data any time after you archive the data. When you repartition data, you use one table column as the partition key. The partition key column can include any datatype except for LOB datatypes. The Data Vault uses the range partitioning method to repartition data.

Users can still query the tables that you repartition. The Data Vault simultaneously manages repartition requests and query requests for the same table. The repartition process occurs in one database commit. Queries only use the repartitioned data files after entire repartition process completes.

When you repartition data, Data Vault performance is not affected. You can use multiple threads to increase the data repartitioning performance.