Replacing a Spool Data Set

If a permanent I/O error occurs on a spool data set and you cannot recover the data (for example, there is a head crash on a direct access device), you can replace the affected spool data set. To replace the data set, perform a warm start and follow the procedures outlined below. You may create the new data set on a volume or device type different from the one being replaced. You may also change the size of the data set and redefine the single track table (STT) range using the STT or STTL parameter on the TRACK or FORMAT initialization statement.

Be aware that when you replace a spool data set, JES3 cancels all jobs with data on the replaced spool data set. Other risks include the possible loss of JES3 control blocks, STT extents, checkpoint records, and the JESNEWS data set, which may have been on the damaged spool data set. If these losses occur, the system will issue messages giving you the opportunity to take appropriate actions.

If you cannot immediately perform a warm start (for example, if it takes some time for you to make the changes needed to replace the spool data set), you can cancel jobs that have track groups allocated on the spool data set being replaced. To cancel the jobs, issue the command *F,Q,DD=ddname,CANCEL. After you cancel the jobs, the user can resubmit them. You can then replace the spool data set at the time most convenient for your installation.

When you replace the spool data set, you must use the same ddname for the new spool data set as for the old.

To replace a spool data set, use the following procedure:
  1. If you allocated the old spool data set by using a JES3 cataloged procedure, update the DD statement in the cataloged procedure to reflect information about the new data set. You may need to change the data set name, device number, device type, or volume serial number. Do not change the ddname.

    If you allocated the old spool data set by including a DYNALLOC statement in the initialization stream, update the optional parameters as necessary. Do not change the ddname.

  2. If the old spool data set is cataloged, replace its catalog entry with an entry for the new spool data set.
  3. If the new spool data set is unformatted and your initialization stream currently includes a TRACK statement for the old spool data set, replace it with a FORMAT statement. Otherwise, leave your TRACK or FORMAT statement alone.
  4. Perform a warm start. Specify WR or WAR as the restart mode. JES3 will prompt you to enter the ddnames of replaced spool data sets (message IAT4009 for unformatted spool data sets and message IAT4008 for formatted spool data sets). JES3 will then cancel all jobs that have track groups allocated to the spool data sets being replaced.