# US 9,519,869 B2 · Predictive computer system resource monitoring

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 9,519,869 B2
- Document ID: US9519869B2
- Status: Granted
- Kind: B2
- Application number: 14/089,098
- Filed: 2013-11-25
- Published: 2015-05-28
- Granted: 2016-12-13
- Earliest claimed: 2013-11-25
- Assignees: International Business Machines Corporation
- Applicants: International Business Machines Corporation
- Inventors: Jinwoo Hwang
- Web page: https://patentgrep.com/patent/US9519869B2

## CPC

- G06F 11/3409: {for performance assessment}
- G06N 20/00: Machine learning
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

A method, system and computer-usable medium are disclosed for monitoring a computer system to predict failed or degraded operational states and respond with an alarm or corrective action. Resource collection and consumption are analyzed to derive velocity and acceleration. A hidden Markov model with the resource collection and consumption data as observation spaces predicts computer system state spaced indicative of a failed or degraded computer system operating state.

## Claims

### Claim 1

A method for monitoring computer system operations, the method comprising:
tracking resource collection at the computer system over time;
tracking resource consumption at the computer system over time;
deriving resource collection velocity and acceleration from the resource collection;
deriving resource consumption velocity and acceleration from the resource consumption;
applying the resource consumption, resource consumption velocity, resource consumption acceleration, resource collection, resource collection velocity and resource collection acceleration as observation spaces in a hidden Markov model to predict computer system state spaces;
comparing the computer system states with an alarm threshold; and
issuing an alarm if the computer system state meets the alarm threshold.

### Claim 2 (depends on 1)

The method of claim 1, further comprising:
comparing the computer system states with a proactive response threshold; and
performing a proactive response if the computer system state meets the proactive response threshold.

### Claim 3 (depends on 1)

The method of claim 1, wherein resource collection comprises assignments of processor and memory resources to a task executing on the computer system.

### Claim 4 (depends on 1)

The method of claim 1, wherein resource consumption comprises use of processor and memory resources by a task executing on the computer system.

### Claim 5 (depends on 1)

The method of claim 1, wherein the computer system state spaces comprise a normal state, a resource contention state, a hang state, a performance degradation state and a crash state.

### Claim 6

A system comprising:
a processor;
a data bus coupled to the processor; and
a computer-usable medium embodying computer program code, the computer-usable medium being coupled to the data bus, the computer program code used for monitoring computer system operations and comprising instructions executable by the processor and configured for:
tracking resource collection at the computer system over time;
tracking resource consumption at the computer system over time;
deriving resource collection velocity and acceleration from the resource collection;
deriving resource consumption velocity and acceleration from the resource consumption;
applying the resource consumption, resource consumption velocity, resource consumption acceleration, resource collection, resource collection velocity and resource collection acceleration as observation spaces in a hidden Markov model to predict computer system state spaces;
comparing the computer system states with an alarm threshold; and
issuing an alarm if the computer system state meets the alarm threshold.

### Claim 7 (depends on 6)

The system of claim 6, wherein the instructions further comprise instructions configured for:
comparing the computer system states with a proactive response threshold; and
performing a proactive response if the computer system state meets the proactive response threshold.

### Claim 8 (depends on 6)

The system of claim 6, wherein resource collection comprises assignments of processor and memory resources to a task executing on the computer system.

### Claim 9 (depends on 6)

The system of claim 6, wherein resource consumption comprises use of processor and memory resources by a task executing on the computer system.

### Claim 10 (depends on 6)

The system of claim 6, wherein the computer system state spaces comprise a normal state, a resource contention state, a hang state, a performance degradation state and a crash state.

### Claim 11

A non-transitory, computer-readable storage medium embodying computer program code, the computer program code comprising computer executable instructions configured for:
tracking resource consumption at the computer system over time;
tracking resource collection at the computer system over time;
deriving resource consumption velocity and acceleration from the resource consumption;
deriving resource collection velocity and acceleration from the resource collection; and
applying the resource consumption, resource consumption velocity, resource consumption acceleration, the resource collection, resource collection velocity and resource collection acceleration as observation spaces in a hidden Markov model to predict computer system state spaces;
adjusting the predicted computer system state spaces with real time data by unsupervised learning with a Viterbi algorithm;
comparing the computer system states with an alarm threshold; and
issuing an alarm if the computer system state meets the alarm threshold.

### Claim 12 (depends on 11)

The non-transitory, computer-readable storage medium of claim 11, wherein the instructions further comprise instructions configured for:
comparing the computer system states with a proactive response threshold; and
performing a proactive response if the computer system state meets the proactive response threshold.

### Claim 13 (depends on 11)

The non-transitory, computer-readable storage medium of claim 11, wherein resource collection comprises assignments of processor and memory resources to a task executing on the computer system.

### Claim 14 (depends on 11)

The non-transitory, computer-readable storage medium of claim 11, wherein resource consumption comprises use of processor and memory resources by a task executing on the computer system.

### Claim 15 (depends on 11)

The non-transitory, computer-readable storage medium of claim 11, wherein the computer executable instructions are deployable to a client system from a server system at a remote location.

### Claim 16 (depends on 11)

The non-transitory, computer-readable storage medium of claim 11, wherein the computer system state spaces comprise a normal state, a resource contention state, a hang state, a performance degradation state and a crash state.