HIPLEX MSCF terminates automatically in the event of irrecoverable errors. The entire HIPLEX MSCF functionality on the processor is cleared down and the processor is disconnected from an XCS network. All imported shared pubsets are exported. The MSCF connections are cleared down and the MSCF subsystem is terminated.
Normally the processor can release all its shared resources. It is possible, however, that tasks might hold up the release of XCS locks (can be displayed via the SHOW-XCS-OCCUPATION command) or the export of shared pubsets (can be displayed via the command SHOW-PUBSET-PARAMETER <catid>,INFORMATION=*PUBSET-OCCUPATION). But MSCF cannot be unloaded until these tasks have released their shared resources. For this reason it might be necessary to terminate the tasks.
A "global crash" occurs if the processor cannot be disconnected from the network due to an error in a registered function or in configuration management. In this case, MSCF can no longer manage the global resources on the processor or cannot terminate access to the global resources, which is why the system must be terminated. If the general configuration parameter RECOVERY-START is set to *AUTOMATIC, an abnormal system termination is initiated (message MCS1301 is output); if *BY-OPERATOR or *CONSISTENT-BY-OPERATOR is set, systems support is requested to shut down the system (message MCS1102 is output).