JES2 isolates bad track groups encountered on JES2 spool volumes. When an unrecoverable I/O error occurs, the operator receives message $HASP094 that identifies the address of the track where the error occurred. During purge processing, track groups that are allocated to the job or data set are normally returned to the pool of available track groups; the defective track groups are not returned. To return a defective track group to service, JES2 attempts to recover that track group. If the recovery is successful for all the tracks on a track group, that track group will be returned to service. The verification takes place when the job that encountered the I/O error is purged and during a warm start. To cause JES2 to remember bad tracks across a cold start, identify them using the BADTRACK initialization statement(s).