Performance Considerations

Encryption is a resource-intensive operation that may introduce a performance overhead to your application in terms of CPU usage and the elapsed time of commands that use encrypted columns.

The amount of overhead depends on the number of CPUs and SAP ASE engines, the load on the system, the number of concurrent sessions accessing the encrypted data, and the number of encrypted columns referenced in a query. The encryption key size and the length of the encrypted data are also factors. In general, the larger the key size and the wider the data, the higher the CPU usage in the encryption operation.

The elapsed time depends on whether the SAP ASE optimizer can make use of an encrypted column.

• Indexes on Encrypted Columns
  You can create an index on an encrypted column if the column’s encryption key does not specify the use of an initialization vector or random padding.
• Sort Orders and Encrypted Columns
  If you use a case-insensitive sort order, SAP ASE cannot use an index on an encrypted char or varchar column when performing a join with another column or a search based on a constant value. This is also true of an accent-insensitive sort order.
• Joins on Encrypted Columns
  SAP ASE optimizes the joining of two encrypted columns by performing cipher text comparisons under certain circumstances.
• Search Arguments and Encrypted Columns
  For equality and nonequality comparison of an encrypted column to a constant value, SAP ASE optimizes the column scan by encrypting the constant value once, rather than decrypting the encrypted column for each row of the table.
• Movement of Encrypted Data as Cipher Text
  As much as possible, SAP ASE optimizes the copying of encrypted data by copying cipher text instead of decrypting and reencrypting data. This applies to select into commands, bulk copying, and replication.