222 Automated Physical Database Design and Tuning
large data warehouses), improving query workload performance by reducing
the I/O costs, and reducing manageability costs by decreasing the time and
storage needed by applications like backup/restore.
While data compression yields significant benefits in the form of reduced
storage costs and reduced I/O, there is a substantial central processing unit
(CPU) cost to be paid in decompressing the data during query processing.
Thus, the decision on when to use compression needs to be made carefully.
Given that compression increases the space of physical design options, there
is a natural motivation to extend automated physical design tools to handle
compression. In order to meet the storage bound as well as ...