# US 2017/0286209 A1 · Techniques for dynamic selection of solutions to storage cluster system trouble events

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

- Number: US 2017/0286209 A1
- Document ID: US20170286209A1
- Status: Published application
- Kind: A1
- Application number: 15/087,475
- Filed: 2016-03-31
- Published: 2017-10-05
- Earliest claimed: 2016-03-31
- Assignees: NetApp, Inc.
- Applicants: Netapp, Inc.
- Inventors: Srirang Heroor; Michael Antony Joseph; Yichao Xue
- Web page: https://patentgrep.com/patent/US20170286209A1

## CPC

- 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)}
- G06F 11/0727: {in a storage system, e.g. in a DASD or network based storage system (drivers for digital recording or reproducing units G06F3/06; circuits for error detection or correction within digital recording or reproducing units G11B20/18; for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS], H04L67/1097)}
- G06F 11/0787: {Storage of error reports, e.g. persistent data storage, storage using memory protection}
- G06F 11/079: {Root cause analysis, i.e. error or fault diagnosis (in a hardware test environment G06F11/22; in a software test environment G06F11/36)}
- G06F 11/3447: {Performance evaluation by modeling}
- G06F 11/3442: {for planning or managing the needed capacity}
- G06F 11/3457: {Performance evaluation by simulation}
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

Various embodiments are generally directed to techniques for dynamic diagnosis and/or prediction of trouble events in a storage cluster system and automated selection of solutions thereto. An apparatus includes a retrieval component to, in response to a trouble event with a first component of a storage cluster system at a usage level under a first usage type, retrieve a component model of a second component associated with a second usage type from a model database, wherein the second usage type comprises operations that differ from operations of the first usage type by no more than a predetermined threshold of difference; and a selection component to apply the first usage level to the component model to derive a resulting level of performance and determine whether to recommend installation of the second component in the first storage cluster system to address the trouble event based on the resulting level of performance.

## Claims

### Claim 1

An apparatus comprising:
a processor component;
a retrieval component to, in response to an indication of a trouble event associated with a first component of a first storage cluster system at a first usage level under a first usage type, retrieve a component model of a second component associated with a second usage type from a model database comprising a plurality of component models of a plurality of components operated under a plurality of usage types, wherein the second usage type comprises a combination of operations that differs from a combination of operations of the first usage type by no more than a first threshold of difference; and
a selection component to apply the first usage level to the component model to derive a resulting level of performance, and to determine whether to provide a recommendation to install the second component in the first storage cluster system to address the trouble event based on the resulting level of performance.

### Claim 2 (depends on 1)

The apparatus of claim 1, the indication of the trouble event comprising a prediction that the trouble event is to occur during operation of the first storage cluster system at the first usage level under the first usage type, the retrieval component to retrieve indications of a recent history of usage levels of the first storage cluster system under the first usage type, and the apparatus comprising a prediction component to predict a highest usage level of the first storage cluster system under the first usage type to occur within a predetermined future time frame as the first usage level in response to an indication in the recent history of a trend of increasing usage of the first storage cluster system under the first usage type.

### Claim 3 (depends on 1)

The apparatus of claim 1, the trouble event comprising at least one of an instance of a level of utilization of a resource provided by the first component that exceeds a predetermined maximum level of utilization, an instance of latency in propagation of data through at least the first component that exceeds a predetermined maximum level of latency, or an instance of a level of data throughput that falls below a predetermined minimum level of data throughput.

### Claim 4 (depends on 3)

The apparatus of claim 3, the selection component to determine whether to apply the first usage level to the component model based on whether an occurrence of the trouble event comprises at least one of the level of utilization of the resource exceeding the predetermined maximum level by a first predetermined degree, the latency in propagation of data through at least the first component exceeding the predetermined maximum level of latency by at least a second predetermined degree, or the level of data throughput falling below the predetermined minimum level of data throughput by at least a third predetermined degree.

### Claim 5 (depends on 3)

The apparatus of claim 3, the selection component to determine whether to provide the recommendation to install the second component in the first storage cluster system based on whether application of the first usage level to the component model results in the level of performance level comprising a resulting level of utilization of the resource that remains below the predetermined maximum level of utilization, a resulting latency in propagation of data through at least the first component that remains below the predetermined maximum level of latency, and a resulting level of data throughput that remains above the predetermined minimum level of data throughput.

### Claim 6 (depends on 1)

The apparatus of claim 1, the selection component to retrieve a list of components that are available to an operator of the first storage cluster system and that are not installed in the first storage cluster system, and the retrieval component to retrieve the component model based in response to inclusion of the second component in the list of components.

### Claim 7 (depends on 1)

The apparatus of claim 1, the retrieval component to retrieve a data point comprising an indication of a second level of usage of the second component in a second storage cluster system under the second usage type, the apparatus comprising an augmenting component to determine whether a combination the second usage type and a component configuration of the second component differs from all other combinations of usage type and component configuration associated with the second component that are included in the model database by at least a second threshold of difference, generate a component model entry in the model database for the combination of the second usage type and component configuration of the second component based on the determination of difference by at least the second threshold of difference, and generate and store the component model in the component model entry in response to receiving a predetermined minimum quantity of data points associated with use of the second component with the component configuration under the second usage type.

### Claim 8

A computer-implemented method comprising:
in response to an indication of a trouble event associated with a first component of a first storage cluster system at a first usage level under a first usage type, retrieving a component model of a second component associated with a second usage type from a model database comprising a plurality of component models of a plurality of components operated under a plurality of usage types, wherein the second usage type comprises a combination of operations that differs from a combination of operations of the first usage type by no more than a first threshold of difference;
applying the first usage level to the component model to derive a resulting level of performance;
determining whether to provide a recommendation to install the second component in the first storage cluster system to address the trouble event based on the resulting level of performance; and
transmitting the recommendation to an administration device for presentation through a network in response to the determination.

### Claim 9 (depends on 8)

The computer-implemented method of claim 8, the indication of the trouble event comprising a prediction that the trouble event is to occur during operation of the first storage cluster system at the first usage level under the first usage type, the method comprising:
retrieving indications of a recent history of usage levels of the first storage cluster system under the first usage type; and
predicting a highest usage level of the first storage cluster system under the first usage type to occur within a predetermined future time frame as the first usage level in response to an indication in the recent history of a trend of increasing usage of the first storage cluster system under the first usage type.

### Claim 10 (depends on 8)

The computer-implemented method of claim 8, the trouble event comprising at least one of an instance of a level of utilization of a resource provided by the first component that exceeds a predetermined maximum level of utilization, an instance of latency in propagation of data through at least the first component that exceeds a predetermined maximum level of latency, or an instance of a level of data throughput that falls below a predetermined minimum level of data throughput.

### Claim 11 (depends on 10)

The computer-implemented method of claim 10, the method comprising determining whether to apply the first usage level to the component model based on whether an occurrence of the trouble event comprises at least one of the level of utilization of the resource exceeding the predetermined maximum level by a first predetermined degree, the latency in propagation of data through at least the first component exceeding the predetermined maximum level of latency by at least a second predetermined degree, or the level of data throughput falling below the predetermined minimum level of data throughput by at least a third predetermined degree.

### Claim 12 (depends on 10)

The computer-implemented method of claim 10, the method comprising determining whether to provide the recommendation to install the second component in the first storage cluster system based on whether application of the first usage level to the component model results in the level of performance level comprising a resulting level of utilization of the resource that remains below the predetermined maximum level of utilization, a resulting latency in propagation of data through at least the first component that remains below the predetermined maximum level of latency, and a resulting level of data throughput that remains above the predetermined minimum level of data throughput.

### Claim 13 (depends on 8)

The computer-implemented method of claim 8, comprising:
retrieving a list of components that are available to an operator of the first storage cluster system and that are not installed in the first storage cluster system; and
selecting the component model for retrieval from the model database in response to inclusion of the second component in the list of components.

### Claim 14 (depends on 8)

The computer-implemented method of claim 8, comprising:
retrieving a data point comprising an indication of a second level of usage of the second component in a second storage cluster system under the second usage type;
determining whether a combination the second usage type and a component configuration of the second component differs from all other combinations of usage type and component configuration associated with the second component that are included in the model database by at least a second threshold of difference;
generating a component model entry in the model database for the combination of the second usage type and component configuration of the second component based on the determination of difference by at least the second threshold of difference; and
generating and storing the component model in the component model entry in response to receiving a predetermined minimum quantity of data points associated with use of the second component with the component configuration under the second usage type.

### Claim 15

At least one non-transitory machine-readable storage medium comprising instructions that when executed by a processor component, cause the processor component to:
in response to an indication of a trouble event associated with a first component of a first storage cluster system at a first usage level under a first usage type, retrieve a component model of a second component associated with a second usage type from a model database comprising a plurality of component models of a plurality of components operated under a plurality of usage types, wherein the second usage type comprises a combination of operations that differs from a combination of operations of the first usage type by no more than a first threshold of difference;
apply the first usage level to the component model to derive a resulting level of performance;
determine whether to provide a recommendation to install the second component in the first storage cluster system to address the trouble event based on the resulting level of performance; and
transmit the recommendation to an administration device for presentation through a network in response to the determination.

### Claim 16 (depends on 15)

The at least one non-transitory machine-readable storage medium of claim 15, the processor component caused to:
retrieve indications of a recent history of usage levels of the first storage cluster system under the first usage type; and
predict a highest usage level of the first storage cluster system under the first usage type to occur within a predetermined future time frame as the first usage level in response to an indication in the recent history of a trend of increasing usage of the first storage cluster system under the first usage type.

### Claim 17 (depends on 15)

The at least one non-transitory machine-readable storage medium of claim 15, the trouble event comprising at least one of an instance of a level of utilization of a resource provided by the first component that exceeds a predetermined maximum level of utilization, an instance of latency in propagation of data through at least the first component that exceeds a predetermined maximum level of latency, or an instance of a level of data throughput that falls below a predetermined minimum level of data throughput.

### Claim 18 (depends on 17)

The at least one non-transitory machine-readable storage medium of claim 17, the processor component caused to determine whether to apply the first usage level to the component model based on whether an occurrence of the trouble event comprises at least one of the level of utilization of the resource exceeding the predetermined maximum level by a first predetermined degree, the latency in propagation of data through at least the first component exceeding the predetermined maximum level of latency by at least a second predetermined degree, or the level of data throughput falling below the predetermined minimum level of data throughput by at least a third predetermined degree.

### Claim 19 (depends on 17)

The at least one non-transitory machine-readable storage medium of claim 17, the processor component caused to determine whether to provide the recommendation to install the second component in the first storage cluster system based on whether application of the first usage level to the component model results in the level of performance level comprising a resulting level of utilization of the resource that remains below the predetermined maximum level of utilization, a resulting latency in propagation of data through at least the first component that remains below the predetermined maximum level of latency, and a resulting level of data throughput that remains above the predetermined minimum level of data throughput.

### Claim 20 (depends on 15)

The at least one non-transitory machine-readable storage medium of claim 15, the processor component caused to:
retrieving a list of components that are available to an operator of the first storage cluster system and that are not installed in the first storage cluster system; and
selecting the component model for retrieval from the model database in response to inclusion of the second component in the list of components.
