Method and apparatus for diagnosing mass storage device anomalies

• US 20070050664A1
• Filed: 08/31/2005
• Published: 03/01/2007
• Est. Priority Date: 08/31/2005
• Status: Active Grant

• Current Assignee: Network Appliance Incorporated (NetApp, Inc.)
• Original Assignee: Network Appliance Incorporated (NetApp, Inc.)
• Inventors: Tan, Cheng-Ping, Phan, Con, Hindi, Faris
• Granted Patent: US 7,461,298 B2
• Time in Patent Office: Days
• US Class Current: 714/5
• CPC Class Codes: 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
• CPC Class Codes: G06F 11/0751 Error or fault detection not based on redundancy power supply failures G06F1/30; network fault management H04L41/06
• CPC Class Codes: G11B 20/1816 Testing
• CPC Class Codes: G11B 2220/2516 Hard disks

Abstract

A type of flaw present in a mass storage device can be inferred by examining the results of I/O operations performed on only a portion of the device, without testing or examining the entire device.

Claims

1. A method comprising:

selecting a first portion of a storage available in a mass storage device, the first portion encompassing less than all of the storage available in the mass storage device;

performing a plurality of I/O operations on the first portion to obtain a corresponding first plurality of I/O results; and

inferring a type of flaw present in the mass storage device based on the first plurality of I/O results.

2. The method of claim 1, wherein the type of flaw is one of transient error, local media damage, storage device failure, and possible head or media failure.

3. The method of claim 1 wherein a size of the first portion of the storage is approximately equal to

4. The method of claim 1, further comprising preparing a defect map of errors detected while performing the plurality of I/O operations and identifying a range containing a significant concentration of errors.

5. The method of claim 4 wherein a span of the range containing a significant concentration of errors is divided by a size of the first portion of storage, the resulting fraction being used to infer a type of flaw present in the mass storage device.

6. The method of claim 1 wherein the first plurality of I/O results indicates more than one possible type of flaw present in the mass storage device, the method further comprising:

selecting a second portion of the storage available in the mass storage device, the second portion not co-extant with the first portion;

performing a plurality of I/O operations on the second portion to obtain a corresponding second plurality of I/O results; and

distinguishing among the more than one possible type of flaw present in the mass storage device based on the second plurality of I/O results.

7. The method of claim 5 wherein a fraction between approximately 0.8/n and approximately 1.2/n, where n is a number of surfaces, and a number of errors not exceeding approximately 20% of a span of the significant error range, indicates two different flaws;

the method further comprising;

selecting a second portion of the storage available in the mass storage device;

performing a second plurality of I/O operations on the second portion to obtain a corresponding second plurality of I/O results; and

comparing the defect map for the first plurality of I/O operations with a defect map for the second plurality of I/O operations.

8. The method of claim 1 wherein the mass storage device is a hard disk drive.

9. A method comprising:

detecting an anomalous condition occurring during a data transfer to or from a mass storage device;

identifying a zone near a location of the data transfer;

executing a plurality of data transfers from the zone;

analyzing a plurality of results obtained from the plurality of data transfers; and

determining a likely cause of the anomalous condition based upon the plurality of results.

10. The method of claim 9 wherein the zone contains less than all of the data contained in the mass storage device.

11. The method of claim 9 wherein analyzing is calculating a fraction equal to a span of a significant number of errors detected during the plurality of data transfers divided by a total number of data transfers.

12. The method of claim 11, further comprising:

when the fraction is near 1/n for n equal to a number of surfaces of the mass storage device,

identifying a second zone far from the location of the data transfer;

executing a second plurality of data transfers from the second zone;

analyzing a second plurality of results obtained from the second plurality of data transfers; and

distinguishing between a first likely cause of the anomalous condition and a second likely cause of the anomalous condition based on the second plurality of results.

13. A system comprising:

a programmable processor;

a plurality of mass storage devices coupled to the processor; and

a sequence of instructions to be executed by the processor, the instructions to;

detect an unusual condition of one of the plurality of mass storage devices;

identify a location associated with the unusual condition;

define a neighborhood near the location;

perform data transfers from the neighborhood; and

diagnose a cause of the unusual condition based upon a plurality of results obtained from the data transfers from the neighborhood.

14. The system of claim 13 wherein the unusual condition is a delayed response to an I/O request.

15. The system of claim 13 wherein the unusual condition is an abnormal status code received from the one of the plurality of mass storage devices.

16. The system of claim 13, further comprising a communication interface to exchange data with a client.

17. The system of claim 16 wherein a function of the system is to provide data storage services to the client.

18. A computer-readable medium containing instructions that, when executed by a processor, cause the processor to perform operations comprising:

detecting an abnormal condition occurring in connection with an I/O operation on a mass storage device;

selecting a portion of the mass storage device centered on a location accessed by the I/O operation;

performing test input operations on the portion of the mass storage device;

collecting a plurality of results from the test operations; and

diagnosing a possible fault in the mass storage device based on the plurality of results.

19. The computer-readable medium of claim 18, wherein diagnosing comprises:

computing a fraction equal to a span of a significant error range divided by a number of sectors of the portion of the mass storage device; and

, if the fraction is near 1/n, for n equal to a number of surfaces of the mass storage device,

selecting a second portion of the mass storage device not overlapping the first portion;

performing additional test input operations on the second portion;

collecting a second plurality of results from the additional test operations; and

using the second plurality of results to refine a diagnosis of the possible fault in the mass storage device.