Replacing damaged RECON data sets

If an I/O error occurs in the current RECON data set that is the only one remaining, DBRC stops the job. Any other jobs that are currently using the RECON data set continue to run if no other I/O error is encountered.

If an I/O error occurs in a RECON data set and two RECON data sets exist, DBRC attempts to locate a spare data set. If a spare is available, DBRC copies the RECON data set without the I/O error to the spare RECON data set. DBRC then establishes the spare as the Copy2 RECON data set.

Recommendation: After the spare RECON data set replaces the RECON data set that had the error, redefine the discarded RECON data set as quickly as possible. If you immediately replace the RECON with the I/O error, you are unlikely to experience a subsystem failure due to loss of all RECON data sets. See Replacing a discarded RECON data set for more information.

If DBRC cannot locate a spare RECON data set and you have specified the STARTNEW parameter of the INIT.RECON command, DBRC continues processing with one RECON data set. Otherwise, DBRC completes the current job but does not start new jobs until you define a spare RECON data set.