Multi-CPU failure detection/recovery system and method for the same

• US 8,074,123 B2
• Filed: 08/20/2009
• Issued: 12/06/2011
• Est. Priority Date: 03/12/2007
• Status: Expired due to Fees

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: Ohira, Yoshiyuki
• Primary Examiner(s): LOHN, JOSHUA A
• Application Number:
  US12/544,618
• Publication Number:
  US 20090307526A1
• Time in Patent Office: 838 Days
• Field of Search: 714/15, 714/16, 714/18, 714/19, 714/49, 714/51, 714/55
• US Class Current: 714/55
• CPC Class Codes: G06F 11/0724 in a multiprocessor or a multi-core unit multiprocessors per se G06F15/80
• CPC Class Codes: G06F 11/0793 Remedial or corrective actions recovery from an exception in an instruction pipeline G06F9/3861; by retry G06F11/1402; for recovering from a failure of a protocol instance or entity H04L69/40

Abstract

A multi-CPU system including plural CPUs, comprising a failure state detection unit for detecting a failure in an operating program, and a recovery unit for determining, when the failure state detection unit has detected a failure, whether or not recovery of data involved in the failure is possible on the basis of content of the detected failure, and for recovering the data when recovery is determined to be possible.

Claims

1. A multi-processor system including processors, comprising:

a failure state detection unit to detect a failure in an operating program and an interruption failure; and

a recovery unit to determine, when the failure state detection unit has detected a failure, whether or not recovery of data involved in the failure is possible on the basis of content of the detected failure, and to recover the data when recovery is determined to be possible,

wherein

when the failure is detected, a processor in which the operating program is running requests a processor other than a processor in which the failure is detected to perform a recovery process; and

an interruption failure is detected by comparing a time that has elapsed since exclusive control was obtained with a measured maximum value of the interruption prohibition time obtained and by comparing a time that has elapsed since obtainment of the exclusive control failed first with the measured maximum value of the interruption prohibition time obtained.

2. The system according to claim 1, wherein the failure state detection unit is to detect a failure in obtainment of exclusive control, and when a time period that has elapsed since obtainment of exclusive control equal or exceeds a specified value;

the operating program is determined to include a failure.

3. The system according to claim 2, wherein the failure state detection is to detect data used for generating the specified value when obtaining and cancelling the exclusive control.

4. The system according to claim 1, wherein the failure state detection unit is to detect an interruption failure as the detected failure, when a time period that has elapsed since interruption prohibition started equals or exceeds a specified value, the operating program is determined to involve a failure.

5. The system according to claim 4, comprising a collection unit to collect data used for generating the specified value when setting and cancelling the interruption prohibition.

6. The system according to claim 1, wherein data used for recovering data is prepared in advance, and when recovery of the data is possible by using the data prepared in advance for recovery of the data, the data is recovered.

7. The system according to claim 1, wherein the recovery unit is to extract, from the data, data that needs to be recovered, and by recovering the extracted data by using the data used for recovery.

8. The system according to claim 1, wherein a program used for recovering data is prepared in advance, and when recovery of the data is possible by using the program prepared in advance for recovery of the data, the data is recovered.

9. A method of detecting a failure and automatic recovery in a multi-processor system including processors, comprising:

detecting a failure in an operating program and an interruption failure;

determining, when the failure state detection step has detected a failure, whether or not recovery of data involved in the failure is possible on the basis of content of the detected failure, and

recovering the data when recovery is determined to be possible,

wherein

when the failure is detected, a processor in which the operating program is running requests a processor other than a processor in which the failure is detected to perform a recovery process; and

an interruption failure is detected by comparing a time that has elapsed since exclusive control was obtained with a measured maximum value of the interruption prohibition time obtained and by comparing a time that has elapsed since obtainment of the exclusive control failed first with the measured maximum value of the interruption prohibition time obtained.

10. A computer-readable storage medium storing a program executed in a multi-processor system including processors, causing the multi-processor system to implement:

a function of detecting a failure in an operating program and an interruption failure; and

a function of determining, when the failure state detection function has detected a failure, whether or not recovery of data involved in the failure is possible on the basis of content of the detected failure, and for recovering the data when recovery is determined to be possible,

wherein

when the failure is detected, a processor in which the operating program is running requests a processor other than a processor in which the failure is detected to perform a recovery process; and

an interruption failure is detected by comparing a time that has elapsed since exclusive control was obtained with a measured maximum value of the interruption prohibition time obtained and by comparing a time that has elapsed since obtainment of the exclusive control failed first with the measured maximum value of the interruption prohibition time obtained.

11. The computer-readable storage medium according to claim 10, wherein:

a failure in obtainment of exclusive control is also detected as the detected failure, and when a time period that has elapsed since obtainment of exclusive control equals or exceeds a specified value, the operating program is determined to include a failure.

12. The computer-readable storage medium according to claim 11, comprising:

collecting data used for generating the specified value when obtaining and cancelling the exclusive control.

13. The computer-readable storage medium according to claim 10, comprising:

collecting data used for generating the specified value when setting and cancelling the interruption prohibition.

14. The computer-readable storage medium according to claim 10, wherein:

data used for recovering data is prepared in advance, and when recovery of the data is possible by using the data prepared in advance for recovery of the data, the data is recovered.

15. The computer-readable storage medium according to claim 10, wherein:

the recovery is performed by extracting, from the data, data that needs to be recovered, and by recovering the extracted data by using the data used for recovery.

16. The computer-readable storage medium according to claim 10, wherein:

a program used for recovering data is prepared in advance, and when recovery of the data is possible by using the program prepared in advance for recovery of the data, the data is recovered.

17. The computer-readable storage medium according to claim 16, wherein:

when there is a link between pieces of data to be recovered, the program used for recovering the data recovers the data by tracing the link.

18. The computer-readable storage medium according to claim 10, wherein:

when a result of the determination indicates that recovery is impossible, the system is shut down.