Failsoft processing

Failsoft processing occurs when no data sets in the primary RACF® database are available (RACF is installed but inactive). Failsoft processing degrades system performance and system security.

There are several reasons why failsoft processing might be in effect on your system:

RACF is installed but does not know the name of the primary master data set.
Failures occurred during RACF initialization at IPL time.
An RVARY INACTIVE command was issued, inactivating all data sets in the primary database.

If RACF is enabled for sysplex communication, failsoft processing can also result when:

The system is running in sysplex local mode.
A data set in the RACF database does not reside on a shared device.
A system is running an MVS™ release that does not support the RACF data sharing option and attempts to IPL in data sharing mode.
A system attempting to join an existing IRRXCF00 group is unable to allocate one or more RACF database data sets that are in use by the other systems in the group. The system operator is not prompted for a data set name, and the system joins the group and begins failsoft processing. A system IPL is required to enable RACF processing when the problem is resolved.
RACF encountered an internal error while processing a request on behalf of the RACF sysplex data sharing group.

During failsoft processing, the operator is prompted frequently to grant access to data sets. To avoid this situation, we recommend that you have a backup RACF database so that you can issue the RVARY SWITCH command rather than an RVARY INACTIVE command.

If you cannot avoid failsoft processing, limit access to the system and do not run production work. You can also try using the RACROUTE REQUEST=AUTH and RACROUTE REQUEST=DEFINE installation exits to implement failsoft processing in some other way.

For more information on failsoft, see Failsoft processing.