Method and apparatus for recovering from errors in a Data Processing System.

摘要

A secondary storage facility employing serial communications between drive (30) and controller (20). A dedicated cable (10) of four conductors is used to connect each controller/drive pair. Each conductor carries one undirectional channel. A first channel (16) is used for command messages and data sent from controller to drive. A second carries response messages and data from drive to controller (14). Third and fourth channels (18 and 12) carry real-time control signals from controller to drive and from drive to controller, respectively, indicative of controller and drive state, for use in synchronizing and coordinating transmission. Each of the real-time state signals comprises a repeating bit pattern of predetermined pattern length: each bit in the pattern has a predetermined meaning. A multi-layered protocol is employed for controller-drive communications over the cable. The first level governs the electrical transmission of messages. Framing and data operations occur at a second level. A third level is used for command and response exchanges. The protocol is synchronous and supports variable length, multiple word messages, each bounded by specific beginning and end frames. Specific command/response exchanges are identified. One is for communicating from the drive to the controller certain parametric information about the drive, which the controller must know to be able to use the drive. Another enables a host computer to selectively disable and enable the read-time drive state signals, to allow the host computer to determine all paths for accessing each drive with which it can communicate. Each drive employs multiple error recovery techniques; the controller signals the drive to try such techniques in sequence, according to descending a priori probability of success. The controller does not know or need to know the details of the error recovery procedures.