Recovering an error range

By using the ERROR RANGE option of RECOVER, you can repair pages with reported I/O errors. Db2 maintains a page error range for I/O errors for each data set; pages within the range cannot be accessed. The DISPLAY DATABASE command displays the range.

When recovering an error range, RECOVER:

  1. Locates, allocates, and applies image copies.
  2. Applies changes from the log.

The following RECOVER statement specifies that the utility is to recover any current error range problems for table space TS1: 

RECOVER TABLESPACE DB1.TS1 ERROR RANGE

Recovering an error range is useful when the range is small, relative to the object containing it; otherwise, recovering the entire object is preferable.

Message DSNU086I indicates that I/O errors were detected on a table space and that you need to recover it. Before you attempt to use the ERROR RANGE option of RECOVER, you should run the ICKDSF service utility to correct the disk error. If an I/O error is detected during RECOVER processing, Db2 issues message DSNU538I to identify the affected target tracks are involved. The message provides enough information to run ICKDSF correctly.

In some situations, which are announced by error messages, recovery of only an error range is not possible. In such a situation, recovering the entire object is preferable.

During the recovery of the entire table space or index space, Db2 might still encounter I/O errors that indicate Db2 is still using a bad volume. For user-defined data sets, you should use Access Method Services to delete the data sets and redefine them with the same name on a new volume. If you use Db2 storage groups, you can remove the bad volume from the storage group by using ALTER STOGROUP. If you use DFSMS storage groups, you should also remove the bad volume from the DFSMS storage group.