Recovering a Database: Step-By-Step Instructions

The symptoms of media failure are as variable as the causes.

If only a single block on the disk is bad, your database may appear to function normally for some time after the corruption occurs, unless you are frequently running dbcc commands. If an entire disk or disk controller is bad. SAP ASE marks the database as suspect and displays a warning message. If the disk storing the master database fails, users cannot log in to the server, and users already logged in cannot perform any actions that access the system tables in master.

  1. Get a current log dump of every database on the device.
  2. Examine the space usage of every database on the device.
  3. After you have gathered this information for all databases on the device, drop each database.
  4. Use sp_dropdevice to drop the failed device. See the Reference Manual: Procedures.
  5. Use disk init to initialize the new database devices. See System Administration Guide: Volume 1 > Initializing Database Devices.
  6. Re-create the databases, one at a time.
  7. Load the most recent database dump into each database.
  8. Apply each transaction log dump in the order in which it was created.

The steps that require more detailed information than listed here are discussed in the following sections.