Storage medium and information processing apparatus and method with failure prediction

Patent metadata

Field	Value
Publication number	US-9317394-B2
Application number	US-201414305055-A
Country	US
Kind code	B2
Filing date	Jun 16, 2014
Priority date	Dec 19, 2011
Publication date	Apr 19, 2016
Grant date	Apr 19, 2016

Abstract

A management computer performs a process comprising: a first step and a second step. The first step collects failure-predictive information including information on a plurality of kinds of phenomena related to occurrence of a failure from each of a plurality of computers including an active first computer in a redundant system that are managed by said management computer. The second step calculates, for each individual second computer of one or more second computers of a plurality of second computers associated with said first computer in said redundant system, an evaluation value that indicates the probability of occurrence of a future failure in the individual second computer using said failure-predictive information collected from the individual second computer and said failure-predictive information collected from one or more predetermined computers other than the individual second computer of said plurality of computers.

First claim

What is claimed is:

1. A non-transitory computer readable storage medium that stores a program that makes a management computer perform a process comprising:
collecting failure-predictive information including information on a plurality of kinds of phenomena related to occurrence of a failure from each of a plurality of computers including an active first computer in a redundant system that are managed by said management computer; and
calculating, for each individual second computer of one or more second computers of a plurality of second computers associated with said first computer in said redundant system, an evaluation value that indicates the probability of occurrence of a future failure in the individual second computer using said failure-predictive information collected from the individual second computer and said failure-predictive information collected from one or more predetermined computers other than the individual second computer of said plurality of computers.

2. The storage medium according to claim 1 , wherein said plurality of kinds of phenomena includes a temperature-related phenomenon and a voltage-related phenomenon, and
said evaluation value depends on a plurality of values each of which indicates the tendency of occurrence of failures each of which is associated with a corresponding one of said plurality of kinds of phenomena in the individual second computer according to said failure-predictive information collected from the individual second computer.

3. The storage medium according to claim 2 , wherein a first value of said plurality of values that is associated with said temperature-related phenomenon is based on the length of time for which each of one or more temperature-related abnormalities has lasted in said individual second computer for which said evaluation value is to be calculated, and
a second value of said plurality of values that is associated with said voltage-related phenomenon is based on the length of time for which each of one or more voltage-related abnormalities has lasted in said individual second computer for which said evaluation value is to be calculated.

4. The storage medium according to claim 2 , wherein said evaluation value further depends on one or both of
a first partial evaluation value that indicates the probability of occurrence of a future failure due to a degradation over time in said individual second computer for which said evaluation value is to be calculated, and
a second partial evaluation value that indicates the probability of occurrence of a future failure in a particular time zone in said individual second computer for which said evaluation value is to be calculated.

5. The storage medium according to claim 4 , wherein said first partial evaluation value is based on at least one of
the length of time for which said individual second computer for which said evaluation value is to be calculated has been powered on,
the length of time for which said individual second computer for which said evaluation value is to be calculated has been powered off,
the total of the length of time for which said individual second computer for which said evaluation value is to be calculated has been powered on and the length of time for which said individual second computer for which said evaluation value is to be calculated has been powered off,
the length of time for which each of one or more temperature-related or voltage-related abnormalities has lasted in said individual second computer for which said evaluation value is to be calculated,
the length of time for which each of the one or more temperature-related or voltage-related abnormalities has lasted in said one or more predetermined computers for which said failure-predictive information collected therefrom is used to calculate said evaluation value for said individual second computer for which said evaluation value is to be calculated, and
the length of time for which each of the one or more abnormalities has lasted in one or more of said plurality of computers that are of the same model as said individual second computer for which said evaluation value is to be calculated.

6. The storage medium according to claim 4 , wherein said second partial evaluation value is based on a history that indicates whether or not an abnormality occurred in said particular time zone in a day in the past in said individual second computer for which said evaluation value is to be calculated.

7. The storage medium according to claim 1 , wherein said one or more predetermined computers for which said failure-predictive information collected therefrom is used to calculate said evaluation value for the individual second computer are one or more computers that
share a power supply unit in a chassis or a power supply unit in a rack with said individual second computer,
is mounted at a position adjacent to said individual second computer,
is mounted in the same chassis as said individual second computer, or
is mounted in the same rack as said individual second computer.

8. The storage medium according to claim 7 , wherein said evaluation value for said individual second computer is calculated by placing more importance on said failure-predictive information collected from said individual second computer than on said failure-predictive information collected from said one or more predetermined computers.

9. The storage medium according to claim 1 , wherein said process further comprising selecting, as a second computer that is to replace said first computer, one second computer for which said evaluation value calculated therefrom indicates a probability equal to or lower than a probability indicated by a predetermined criterion, and
said criterion is determined by a predetermined threshold for said evaluation value or concerns a relative order of said evaluation values among said plurality of second computers.

10. An information processing apparatus comprising:
a receiver that receives failure-predictive information including information on a plurality of kinds of phenomena related to occurrence of a failure from each of a plurality of computers including an active first computer in a redundant system that are managed by said information processing apparatus;
a storage unit that stores said failure-predictive information received by said receiver; and
a processor that calculates, for each individual second computer of one or more second computers of a plurality of second computers associated with said first computer in said redundant system, an evaluation value that indicates the probability of occurrence of a future failure in the individual second computer using said failure-predictive information received by said receiver from the individual second computer and stored in said storage unit and said failure-predictive information received by said receiver from one or more predetermined computers other than the individual second computer of said plurality of computers and stored in said storage unit.

11. A method performed by a management computer, the method comprising:
collecting, by a management computer, failure-predictive information including information on a plurality of kinds of phenomena related to occurrence of a failure from each of a plurality of computers including an active first computer in a redundant system that are managed by said management computer, and
calculating, by the management computer, for each individual second computer of one or more second computers of a plurality of second computers associated with said first computer in said redundant system, an evaluation value that indicates the probability of occurrence of a future failure in the individual second computer using sai

Assignees

• Fujitsu Ltd

Inventors

• Sakurai Kazuto
• Sawada Hitoshi
• Tamura Naohiro

Classifications

• G06F11/2025 Primary

  using centralised failover control functionality · CPC title

• G06F11/3452 Primary

  Performance evaluation by statistical analysis · CPC title

• G06F11/2041

  with more than one idle spare processing component · CPC title

• G06F11/2028

  eliminating a faulty processor or activating a spare · CPC title

• G06F2201/86

  Event-based monitoring · CPC title