Storage device management in computing systems

• US 9,785,374 B2
• Filed: 09/25/2014
• Issued: 10/10/2017
• Est. Priority Date: 09/25/2014
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Inventors: Santaniello, Mark, Khessib, Badriddine, Caulfield, Laura, Sharma, Bikash
• Primary Examiner(s): Lazaro, David
• Assistant Examiner(s): SHITAYEWOLDETADIK, BERHANU
• Application Number:
  US14/496,078
• Publication Number:
  US 20160092140A1
• Time in Patent Office: 1,111 Days
• Field of Search: 709224-245
• CPC Class Codes: G06F 12/00 Accessing, addressing or allocating within memory systems or architectures digital input from, or digital output to record carriers, e.g. to disk storage units, G06F3/06
• CPC Class Codes: G06F 3/0611 in relation to response time
• CPC Class Codes: G06F 3/0616 in relation to life time, e.g. increasing Mean Time Between Failures [MTBF]
• CPC Class Codes: G06F 3/0647 Migration mechanisms
• CPC Class Codes: G06F 3/0683 Plurality of storage devices
• CPC Class Codes: G06F 3/0688 Non-volatile semiconductor memory arrays
• CPC Class Codes: G06F 9/5088 involving task migration
• CPC Class Codes: H04L 67/1008 based on parameters of servers, e.g. available memory or workload monitoring of computer activity G06F11/30
• CPC Class Codes: H04L 67/1097 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]

Abstract

Various techniques of managing storage devices in a computing system are described in this application. In one embodiment, a method includes receiving an input containing consumption data representing consumption of a storage device in one of the processing units and determining if the storage device in one of the processing units is consumed excessively. In response to determining that the storage device is consumed excessively, an indicator may be generated to indicate a potential program migration from the one of the processing units to another one of the processing units in the computing system.

Claims

1. A computing device having a processor and a memory containing instructions that when executed by the processor, cause the processor to perform a process comprising:

receiving an input from a first processing unit of a computing system, the first processing unit executing one or more application programs with a processor, wherein the input containing data representing a number of operations previously performed by a storage device of the first processing unit;

determining if the number of operations of the storage device of the first processing unit is excessive in relation to a target threshold; and

in response to determining that the number of operations of the storage device of the first processing unit is excessive, migrating one or more of the application programs from the first processing unit to a second processing unit in the computing system, wherein the process performed by the processor further includes determining the target threshold based on an endurance of the storage device and a time that the storage device is in service.

2. The computing device of claim 1 wherein the one or more application programs includes one or more virtual machines, and wherein migrating one or more of the application programs includes migrating one or more of the virtual machines from the first processing unit to the second processing unit.

3. The computing device of claim 1 wherein the number of operations previously performed by the storage device includes an accumulated number of reads, writes, and/or erases previously committed to the storage device, and wherein determining if the number of operations to the storage device of the first processing unit is excessive includes determining if the number of reads, writes, and/or erases previously committed to the storage device is above the target threshold.

4. The computing device of claim 1 wherein:

the number of operations previously performed by the storage device includes an accumulated number of writes previously committed to the storage device; and

the process performed by the processor further includes;

calculating a percentage of excess in the number of writes previously committed to the storage device based on an expected number of writes; and

comparing the calculated percentage of excess to the target threshold.

5. The computing device of claim 1 wherein:

the number of operations previously performed by the storage device includes an accumulated number of writes previously committed to the storage device; and

the process performed by the processor further includes;

calculating a percentage of excess in the number of writes previously committed to the storage device based on an expected number of writes;

determining if the calculated percentage of excess is larger than the target threshold; and

in response to determining that the percentage of excess is larger than the target threshold, indicating that the number of operations to the storage device of the first processing unit is excessive.

6. The computing device of claim 1 wherein:

the number of operations previously performed by the storage device includes an accumulated number of writes previously committed to the storage device; and

the process performed by the processor further includes;

calculating a percentage of excess in the number of writes previously committed to the storage device based on an expected number of writes;

determining if the calculated percentage of excess is larger than the target threshold; and

in response to determining that the percentage of excess is not larger than the target threshold, performing at least one of indicating that the number of operations to the storage device of the first processing unit is not excessive;

or marking the processing unit as suitable as a migrate target.

7. A method of managing storage devices in a computing system having a plurality of servers interconnected with one another via a computer network, the servers individually having a processor and one or more storage devices operatively coupled to the processor, the method comprising:

receiving, from one of the servers via the computer network, an input containing consumption data representing a consumption status of a storage device coupled to a processor of one of the servers, wherein the storage device is accessed by the processor when executing instructions to host a virtual machine on the one of the servers;

determining whether the storage device coupled to the one of the servers is consumed excessively based on an expected consumption rate of the storage device and the received consumption status of the storage device from the one of the servers, wherein the expected consumption rate is determined based on an endurance of the storage device and a time that the storage device is in service; and

in response to determining that the storage device is consumed excessively, causing the virtual machine currently hosted on the one of the servers to migrate, via the computer network, from the one of the servers to be hosted by another one of the servers in the computing system.

8. The method of claim 7 wherein determining whether the storage device in the one of the processing units is consumed excessively includes:

calculating a consumption variation based on the received consumption data and the expected consumption of the storage device;

determining whether the calculated consumption variation associated with the storage device is above a threshold; and

in response to determining that the calculated consumption variation is above the threshold, indicating that the storage device is consumed excessively.

9. The method of claim 8 wherein calculating the consumption variation of the storage device includes calculating at least one of a difference between the received consumption data and the expected consumption of the storage device or a percentage of excess consumption above the expected consumption of the storage device.

10. The method of claim 8 wherein calculating the consumption variation of the storage device includes calculating the expected consumption based on a life endurance of the storage device and a time in service of the storage device.

11. The method of claim 8 wherein calculating the consumption variation of the storage device includes:

calculating an expected remaining endurance of the storage device by multiplying a life endurance of the storage device and a time in service of the storage device; and

calculating a percentage of excess consumption based on the calculated remaining endurance and the data representing the consumption of the storage device.

12. The method of claim 7 wherein the computing system includes a data center.

13. The method of claim 7, further comprising:

in response to determining that the storage device is consumed excessively, selecting the another one of the servers of the computing system based on a consumption status of a storage device on the another server; and

automatically migrating the virtual machine from the one of the servers to the selected another server via the computer network.

14. The method of claim 7, further comprising:

in response to determining that the storage device is consumed excessively, selecting the another one of the servers of the computing system based at least in part on a physical proximity of the another one of the servers to the one of the servers; and

automatically migrating the virtual machine from the one of the servers to the selected another one of the servers via the computer network.

15. The method of claim 7 wherein:

in response to determining that the storage device is consumed excessively, selecting a storage device coupled to the another one of the servers of the computing system; and

allowing the processor of the one of the servers to access the selected storage device of the another one of the servers via the computer network.

16. A method of managing storage devices in a computing system having a plurality of servers interconnected with one another via a computer network, the servers individually having a processor and a storage device operatively coupled to the processor, the method comprising:

receiving, from a first server of the computing system via the computer network, an input containing consumption data representing a consumption status of a storage device coupled to a first processor of the first server, wherein the storage device is accessible by the first processor when executing instructions of an application on the first server;

determining whether the storage device coupled to the first processor of the first server is consumed excessively based on a comparison between the received consumption data of the storage device and a projected consumption rate of the storage device, wherein the projected consumption rate is determined based on an endurance of the storage device and a time that the storage device is in service; and

in response to determining that the storage device of the first server is consumed excessively, causing the application currently being executed by the first processor of the first server to migrate, via the computer network, from the first server to a second server in the computing system to be executed by a second processor of the second server.

17. The method of claim 16 wherein determining whether the storage device coupled to the first processor of the first server is consumed excessively includes:

calculating an average rate of consumption of all storage devices in the servers of the computing system; and

determining a variation of rate of consumption for the storage device of the first server from the calculated average rate of consumption.

18. The method of claim 16 wherein determining whether the storage device coupled to the first processor of the first server is consumed excessively includes:

calculating an average rate of consumption of all storage devices in the servers of the computing system;

determining a variation of rate of consumption for the storage device of the first server from the calculated average rate of consumption determining whether the variation of rate of consumption is greater than a threshold; and

in response to determining that the variation of rate of consumption is greater than the threshold, indicating that the storage device of the first server is consumed excessively relative to the other storage devices in the computing system.

19. The method of claim 16 wherein determining whether the storage device coupled to the first processor of the first server is consumed excessively includes:

calculating an average rate of consumption of all storage devices in the computing system;

determining a percentage of variation for the storage device of the first server based on the calculated average rate of consumption;

determining whether the percentage of variation is greater than a threshold; and

in response to determining that the percentage of variation is greater than the threshold, indicating that the storage device of the first server is consumed excessively relative to the other storage devices in the computing system.