The structure of the system of automatic control and diagnostics of computers

The structure of the system of automatic control and diagnostics of computers

System definition
Functions and algorithms
General system requirements
Control circuit checker
The structure of the system of automatic monitoring and diagnostics is called an error detection system. The occurrence of an error at any computer causes a signal on which program execution is suspended. Subsequent errors can be different. Therefore, it is necessary to determine the nature and method of the error, if it fails, you need to restore the reliability of information. then a repair notification is received
Control is the determination of the correct operation of the device object.
Diagnostics is a sequential troubleshooting process.

Functions of the structure of the system of automatic control and diagnostics

Recognizing the nature of the error
Re-start the program
Fault location
Record in memory v.n. information on all failure failures for later analyzes

Block diagram of the algorithm of the automatic monitoring and diagnostics

Program
recording the computer's situation at the moment the error
Diagnosing the nature of the error by repeating the program
Device or error
Record Failure Information
Diagnostic procedures

Rollback localized
Possibility or reconstruction

Post Failure Message

Automatic shutdown of a faulty device

Repairs

Work algorithm

Write special registers or memory
The situation at the time of the error
Interrupt signal processing diagnostics by repeating an operation or program

If the error does not repeat. This is a failure and control is presented to the breakpoint.
If the error is repeated during the repetition, the fault finding system starts working.

After localization of the fault location, reconfiguration or repair takes place.

Reconfiguration is the redistribution of available resources.

Requirements

Detecting a certain percentage of faults
Diagnose and localize the location of the error
Detect an error quickly enough to prevent it from spreading in the computational process
Categorize the error To ensure the correct response to it.

Failure failure
Intermittent Malfunction Multiple Failure
Software errors
Operator error
Errors in control circuits

Incorrect operation of control circuits may cause
To her error detection
To the appearance of a false error signal
Such schemes are periodically checked by creating artificial errors.

For example, checking control circuits of information channels checks data with faulty control bits, if it is not possible to check the control circuit, then duplication of the control system is introduced to increase reliability