Performance of object storage system by reconfiguring storage devices based on latency that includes identifying a number of fragments that has a particular storage device as its primary storage device and another number of fragments that has said particular storage device as its replica storage device

• US 10,664,169 B2
• Filed: 06/24/2016
• Issued: 05/26/2020
• Est. Priority Date: 06/24/2016
• Status: Active Grant

• Current Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Original Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Inventors: George, Johnu, Zhang, Kai, Tarre, Marc Solanas, Dutta, Debojyoti, Sharma, Manoj
• Primary Examiner(s): Partridge, William B
• Assistant Examiner(s): Yew, Chie
• Application Number:
  US15/192,255
• Publication Number:
  US 20170371558A1
• Time in Patent Office: 1,432 Days
• Field of Search: None
• CPC Class Codes: G06F 3/0611 in relation to response time
• CPC Class Codes: G06F 3/0631 by allocating resources to storage systems
• CPC Class Codes: G06F 3/0643 Management of files
• CPC Class Codes: G06F 3/0653 Monitoring storage devices or systems
• CPC Class Codes: G06F 3/067 Distributed or networked storage systems, e.g. storage area networks [SAN], network attached storage [NAS]

Abstract

Approaches are disclosed for improving performance of logical disks. A logical disk can comprise several storage devices. In an object storage system (OSS), when a logical disk stores a file, fragments of the file are stored distributed across the storage devices. Each of the fragments of the file is asymmetrically stored in (write) and retrieved from (read) the storage devices. The performance of the logical disk is improved by reconfiguring one or more of the storage devices based on an influence that each of the storage devices has on performance of the logical disk and the asymmetric read and write operations of each of the storage devices. For example, latency of the logical disk can be reduced by reconfiguring one or more of the plurality of storage disks based on a proportion of the latency of the logical device that is attributable to each of the plurality of storage devices.

Claims

1. A method comprising:

storing a file in a logical disk, wherein fragments of the file are stored distributed across a plurality of storage devices comprising the logical disk and each of the fragments of the file is asymmetrically stored in and retrieved from the plurality of storage devices; and

reducing latency of the logical disk based on the asymmetrical storage in and retrieval from the plurality of storage devices by;

calculating, for each of the plurality of storage devices, at least one impact factor that impacts performance, the at least one impact factor identifying a proportion of the fragments stored in or retrieved from each of the plurality of storage devices relative to others of the plurality of storage devices based on the asymmetrical storage in and retrieval from each of the plurality of storage devices;

calculating a proportion of a latency of the logical disk that is attributable to each of the plurality of storage devices based, at least in part, on the at least one impact factor and a latency of each of the plurality of storage devices; and

reconfiguring one or more of the plurality of storage devices based on the proportion of the latency of the logical disk that is attributable to each of the plurality of storage devices,

wherein, the at least one impact factor includes a read impact factor and a write impact factor,

the read impact factor identifies a first number of fragments that has a particular storage device of the plurality of storage devices as its primary storage device,

and the write impact factor identifies a sum of the first number of fragments and a second number of fragments that has the particular storage device as its replica storage device.

2. The method of claim 1, wherein the reconfiguring one or more of the plurality of storage devices based on the proportion of the latency of the logical disk that is attributable to each of the plurality of storage devices comprises:

determining which of the plurality of storage devices is a storage device for removal from the logical disk is based on the proportion of the latency of the logical disk that is attributable to the storage device; and

removing the storage device from the logical disk.

3. The method of claim 2, wherein the removing the storage device from the logical disk comprises:

transferring data from the storage device to others of the plurality of storage devices by pseudo-randomly distributing from the storage device to the others of the plurality of storage devices all fragments associated with files stored in the logical disk;

disassociating the storage device from the logical disk; and

deleting, from the storage device, all the fragments associated with files stored in the logical disk.

4. The method of claim 1, wherein the reconfiguring one or more of the plurality of storage disks based on the proportion of the latency of the logical disk that is attributable to each of the plurality of storage devices comprises:

associating a new storage device to the plurality of storage devices comprising the logical disk; and

transferring data from others of the plurality of storage devices to the new storage device by pseudo-randomly distributing from the others of the plurality of storage devices a portion of fragments associated with files stored in the logical disk.

5. The method of claim 2, wherein the proportion for the storage device is larger than the proportion for others of the plurality of storage devices.

6. The method of claim 1, wherein the each of the fragments of the file being asymmetrically stored in and retrieved from the plurality of storage devices comprises the logical disk accessing the plurality of storage devices a different number of times for retrieving each of the fragments than for writing each of the fragments.

7. The method of claim 1, wherein the each of the fragments of the file being asymmetrically stored in and retrieved from the plurality of storage devices comprises:

writing, by the logical disk, a fragment of the fragments to a first storage device as its primary storage device and to each of one or more second storage devices as its replica storage device, of the plurality of storage devices;

and retrieving, by the logical disk, the fragment from the first storage device and not from the one or more second storage devices.

8. The method of claim 7, wherein the calculating the proportion of the latency of the logical disk that is attributable to each of the plurality of storage devices comprises:

retrieving the latency of each of the plurality of storage devices, wherein the latency of each of the plurality of storage devices is latency of input/output (“

I/O”

) operations measured during a window of time, determining a weighting factor for each of the plurality of storage devices based, at least in part, on the read impact factor and/or the write impact factor; and

determining a weighted latency for each of the plurality of storage devices based on the latency of the I/O operations measured during the window of time and the weighting factor by multiplying the latency of the I/O operations by the weighting factor.

9. The method of claim 8, further comprising:

calculating a latency of the logical disk by summing the weighted latency for each of the plurality of storage devices.

10. The method of claim 1, wherein the at least one impact factor is a number of operations in a queue of operations to be performed by each of the plurality of storage devices, a total number of objects stored by each of the plurality of storage devices, and/or a total number of logical disks using each of the plurality of storage devices.

11. A system comprising:

a logical disk configured to store a file, wherein fragments of the file are stored distributed across a plurality of storage devices comprising the logical disk and each of the fragments of the file is asymmetrically stored in and retrieved from the plurality of storage devices;

and a network element configured to reduce latency of the logical disk based on the asymmetrical storage in and retrieval from the plurality of storage devices by;

calculating, for each of the plurality of storage devices, at least one impact factor that impacts performance, the at least one impact factor identifying a proportion of the fragments stored in or retrieved from each of the plurality of storage devices relative to others of the plurality of storage devices based on the asymmetrical storage in and retrieval from each of the plurality of storage devices;

calculating a proportion of a latency of the logical disk that is attributable to each of the plurality of storage devices based, at least in part, on the at least one impact factor and a latency of each of the plurality of storage devices;

and reconfiguring one or more of the plurality of storage devices based on the proportion of the latency of the logical disk that is attributable to each of the plurality of storage devices,

wherein,

the at least one impact factor includes a read impact factor and a write impact factor,

the read impact factor identifies a first number of fragments that has a particular storage device of the plurality of storage devices as its primary storage device,

and the write impact factor identifies a sum of the first number of fragments and a second number of fragments that has the particular storage device as its replica storage device.

12. The system of claim 11, wherein the each of the fragments of the file being asymmetrically stored in and retrieved from the plurality of storage devices comprises:

writing, by the logical disk, a fragment of the fragments to a first storage device as its primary storage device and to each of one or more second storage devices as its replica storage device, of the plurality of storage devices;

and retrieving, by the logical disk, the fragment from the first storage device and not from the one or more second storage devices.

13. The system of claim 12, wherein the calculating the proportion of the latency of the logical disk that is attributable to each of the plurality of storage devices comprises:

retrieving the latency of each of the plurality of storage devices, wherein the latency of each of the plurality of storage devices is latency of input/output (“

I/O”

) operations measured during a window of time;

determining a weighting factor for each of the plurality of storage devices based, at least in part, on the read impact factor and/or the write impact factor; and

determining a weighted latency for each of the plurality of storage devices based on the latency of the I/O operations measured during the window of time and the weighting factor by multiplying the latency of the I/O operations by the weighting factor.

14. The system of claim 11, wherein the at least one impact factor is a number of operations in a queue of operations to be performed by each of the plurality of storage devices, a total number of objects stored by each of the plurality of storage devices, and/or a total number of logical disks using each of the plurality of storage devices.

15. A computer-readable non-transitory medium comprising instructions, that when executed by at least one processor configure the at least one processor to perform operations comprising:

storing a file in a logical disk, wherein fragments of the file are stored distributed across a plurality of storage devices comprising the logical disk and each of the fragments of the file is asymmetrically stored in and retrieved from the plurality of storage devices; and

reducing latency of the logical disk based on the asymmetrical storage in and retrieval from the plurality of storage devices by;

calculating, for each of the plurality of storage devices, at least one impact factor that impacts performance, the at least one impact factor identifying a proportion of the fragments stored in or retrieved from each of the plurality of storage devices relative to others of the plurality of storage devices based on the asymmetrical storage in and retrieval from each of the plurality of storage devices;

calculating a proportion of a latency of the logical disk that is attributable to each of the plurality of storage devices based, at least in part, on the at least one impact factor and a latency of each of the plurality of storage devices; and

reconfiguring one or more of the plurality of storage devices based on the proportion of the latency of the logical disk that is attributable to each of the plurality of storage devices,

wherein,

the at least one impact factor includes a read impact factor and a write impact factor,

the read impact factor identifies a first number of fragments that has a particular storage device of the plurality of storage devices as its primary storage device,

and the write impact factor identifies a sum of the first number of fragments and a second number of fragments that has the particular storage device as its replica storage device.

16. The computer-readable non-transitory medium of claim 15, wherein the each of the fragments of the file being asymmetrically stored in and retrieved from the plurality of storage devices comprises:

writing, by the logical disk, a fragment of the fragments to a first storage device as its primary storage device and to each of one or more second storage devices as its replica storage device, of the plurality of storage devices;

and retrieving, by the logical disk, the fragment from the first storage device and not from the one or more second storage devices.

17. The computer-readable non-transitory medium of claim 16, wherein the calculating the proportion of the latency of the logical disk that is attributable to each of the plurality of storage devices comprises:

retrieving the latency of each of the plurality of storage devices, wherein the latency of each of the plurality of storage devices is latency of input/output (“

I/O”

) operations measured during a window of time;

determining a weighting factor for each of the plurality of storage devices based, at least in part, on the read impact factor and/or the write impact factor; and

determining a weighted latency for each of the plurality of storage devices based on the latency of the I/O operations measured during the window of time and the weighting factor by multiplying the latency of the I/O operations by the weighting factor.

18. The computer-readable non-transitory medium of claim 15, wherein the at least one impact factor is a number of operations in a queue of operations to be performed by each of the plurality of storage devices, a total number of objects stored by each of the plurality of storage devices, and/or a total number of logical disks using each of the plurality of storage devices.