Error-Correction Coding Scheme To Allow Marking Of Stale Data
Original Publication Date: 1979-Jul-01
Included in the Prior Art Database: 2005-Feb-20
This scheme for adding a stale data marking capability to an existing error-correcting code (ECC) preserves data integrity for both marked and unmarked data. It is specifically applicable to a distributed processing network wherein a pool of data processors share a common memory bank. Each of the processors has its own buffer area to which it has exclusive writing privileges, but which may be read by any of the other processors as part of a computer-to-computer data transfer. Data is considered "stale" when, as a result of hardware failure or initialization of the writing processor, the data is no longer valid. Marking the data, as such, is intended to signal the other processors.