Managing hiperspace storage across a checkpoint/restart operation

A program can use checkpoint/restart while it has one or more entries for a hiperspace on its access list (DU-AL or PASN-AL). If the program has specified on the ALESERV macro that the system is to ignore entries made to the access list for the hiperspace for checkpoint/restart processing (CHKPT=IGNORE), the CHKPT macro processes successfully.

A program that specifies CHKPT=IGNORE assumes full responsibility for managing the hiperspace storage. Managing the hiperspace storage includes the following:

• If any program depends on the contents of the hiperspace and the data cannot be recreated or obtained elsewhere, the responsible program must save the contents of the hiperspace prior to the checkpoint operation.
• Once the checkpoint operation has completed, the responsible program must perform the following during restart processing to successfully manage the hiperspace storage.
  1. Ensure that the hiperspace exists. The original hiperspace might or might not exist. If the original hiperspace does not exist, the responsible program must perform a DSPSERV CREATE TYPE=HIPERSPACE to recreate the hiperspace.
  2. Perform an ALESERV ADD of the hiperspace, original or recreated, to the program's access list to obtain a new ALET.
  3. If, in addition to having a dependency on the hiperspace, any program also depends on the contents of the hiperspace storage, the responsible program must refresh the contents of the hiperspace storage. The program must use the new ALET to reference the hiperspace.
  4. The responsible program must make the new ALET available to any program that has a dependency on the hiperspace. The STOKEN, changed or unchanged, must be made available to any program that needs to perform an ALESERV ADD to access the hiperspace.

See z/OS DFSMSdfp Checkpoint/Restart for more information about the CHKPT macro.