Detection of Channel and Control Unit Hang Ups
Original Publication Date: 1980-Jun-01
Included in the Prior Art Database: 2005-Feb-13
In contemporary data processing systems, malfunctions in the I/O sub-system, due to hang-up conditions in I/O channel or control unit entities, often can not be traced to the specific source. When this occurs, the central system and its supervisory programs must be re-initialized, thereby severely impacting system performance. This article concerns a method for effectively tracing the specific locus of a channel or control unit hang-up condition, and has particular applicability to systems in which plural I/O channel interfaces operate through a shared processor (i.e., a "director") which is alterably microprogrammed.