Critical Failure

When an application makes a protocol error in sending data, the local node rejects the data using a Status-Acknowledge(Nack-2) with a sense code indicating the reason for failure. This message has a critical failure flag that indicates whether the local node has marked the session as unrecoverable. The sense codes are listed in FMI Status, Error, and Sense Codes.

If the error is noncritical, the application can proceed as if the message that caused the error had not been sent. If the error is critical, then the local node issues a Close(PLU) Request to the application (providing that the PLU connection is open), which means that the application cannot communicate on the PLU-SLU session until an UNBIND–BIND sequence is received from the host. The local node also sends a TERM-SELF request to the host to elicit an UNBIND; therefore, the application does not need to issue a LOGOFF request on the SSCP session.