SELECTION OF FABRIC-ATTACHED STORAGE DRIVES ON WHICH TO PROVISION DRIVE VOLUMES FOR REALIZING LOGICAL VOLUME ON CLIENT COMPUTING DEVICE WITHIN STORAGE AREA NETWORK

• US 20180150244A1
• Filed: 11/30/2016
• Published: 05/31/2018
• Est. Priority Date: 11/30/2016
• Status: Active Grant

• Current Assignee: Lenovo Global Technologies International Ltd. (Lenovo Group Ltd.)
• Original Assignee: Lenovo Enterprise Solutions (Singapore) Pte., Ltd. (Lenovo Group Ltd.)
• Inventors: Caporale, Patrick Leo, Condict, Michael Neil, Cosby, David W., Hinkle, Jonathan Randall
• Granted Patent: US 10,353,602 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 3/0608 Saving storage space on storage systems
• CPC Class Codes: G06F 3/0619 in relation to data integrity, e.g. data losses, bit errors
• CPC Class Codes: G06F 3/0631 by allocating resources to storage systems
• CPC Class Codes: G06F 3/065 Replication mechanisms
• CPC Class Codes: G06F 3/067 Distributed or networked storage systems, e.g. storage area networks [SAN], network attached storage [NAS]

Abstract

A storage-area network (SAN) manager running on a computing device connected to a fabric receives a request to allocate and mount a logical volume on a client computing device connected to the fabric. The manager groups storage drives connected to the fabric by failure domains. The manager filters the storage drives by native capabilities, available capacities, and dynamic performance to remove from consideration any storage drive having insufficient native capabilities, an insufficient available capacity, and/or an insufficient dynamic performance to satisfy the request. The manager selects a set of storage drives on which to store data of the logical volume, from the remaining storage drives, taking into account the failure domains. The manager sends a command to each such storage drive to provision a drive volume, and a command to a software agent on the client device to create and mount the logical volume using these drive volumes.

Claims

1. A method comprising:

receiving, by a storage-area network (SAN) manager running on a computing device directly connected to a fabric, a request from an administrator to allocate and mount a logical volume on a client computing device directly connected to the fabric;

grouping, by the SAN manager, storage drives directly connected to the fabric by failure domains in which the storage drives are located;

filtering, by the SAN manager, the storage drives by native capabilities, available capacities, and dynamic performance thereof to remove from selection consideration any storage drive having one or more of insufficient native capabilities, an insufficient available capacity, or an insufficient dynamic performance to satisfy the request;

selecting, by the SAN manager, a set of storage drives on which to store data of the logical volume, from the storage drives that remain for the selection consideration after the storage drives have been filtered, including taking into account the failure domains in which the storage drives of the set are located in satisfaction of the request;

sending, by the SAN manager, a command to each storage drive of the selected set to provision a drive volume; and

sending, by the SAN manager, a command to a SAN software agent on the client computing device to create and mount the logical volume using the drive volumes provisioned on the storage drives of the set.

2. The method of claim 1, wherein each failure domain corresponds to a group of the storage drives such that a failure corresponding to the failure domain causes failure of each storage drive of the group.

3. The method of claim 2, wherein grouping the storage drives by the failure domains comprises:

for each failure sub-domain type of a plurality of failure sub-domain types, grouping the storage drives by failure sub-domains corresponding to the failure sub-domain type; and

grouping the storage drives that are located within a same failure sub-domain for every failure sub-domain type within a same failure domain.

4. The method of claim 2, wherein grouping the storage drives by the failure domains comprises:

preliminarily grouping the storage drives by failure domains in which the storage drives are located;

determining a number of required failure domains from the request; and

combining the failure domains into which the storage drives have been grouped until the storage drives have been grouped into a number of the failure domains equal to the number of required failure domains.

5. The method of claim 4, wherein selecting the set of storage drives comprises selecting each storage drive of the set from a different failure domain.

6. The method of claim 4,

wherein combining the failure domains comprises taking into account redundancy importance across the failure domains.

7. The method of claim 1, wherein the native capabilities of each storage drive comprise one or more of:

a type of the storage drive;

a connection technology of the storage drive;

a rated data transfer rate of the storage drive;

an internal data block storage size of the storage drive;

a storage technology of the storage drive.

8. The method of claim 7, wherein selecting the set of storage drives comprises selecting the storage drives of the set to maximize sameness in the native capabilities thereof.

9. The method of claim 1, wherein selecting the set of storage drives comprises selecting the storage drives of the set to minimize remaining capacity on each storage drive after the drive volume has been provisioned on the storage drive.

10. The method of claim 1, wherein selecting the set of storage drives comprises selecting the storage drives of the set to maximize remaining capacity on each storage drive after the drive volume has been provisioned on the storage drive.

11. The method of claim 1, wherein filtering the storage drives comprises, after filtering the storage drives to remove from the selection consideration any storage drive having one or more of insufficient native capabilities, or the insufficient capacity:

determining the dynamic performance for each storage drive that remains.

12. The method of claim 11, wherein determining the dynamic performance for each storage drive that remains comprises one of:

retrieving from the storage drive dynamic characteristics maintained by the storage drive itself that are indicative of the dynamic performance of the storage drive;

measuring the dynamic performance between the SAN manager and the storage drive;

measuring the dynamic performance between the client computing device and the storage drive.

13. A non-transitory computer-readable data storage medium storing computer-executable code that a computing device directly connected to a fabric and running a storage-area network (SAN) manager executes to perform a method comprising:

receiving a request from an administrator to allocate and mount a logical volume on a client computing device directly connected to the fabric;

selecting a set of storage drives on which to store data of the logical volume, from storage drives directly connected to the fabric, based on native capabilities, available capacities, and dynamic performance of the storage drives, and taking into account failure domains in which the storage drives of the set are located in satisfaction of the request;

sending a command to each storage drive of the selected set to provision a drive volume; and

sending a command to a SAN software agent on the client computing device to create and mount the logical volume using the drive volumes provisioned on the storage drives of the set.

14. The non-transitory computer-readable data storage medium of claim 13, wherein each failure domain corresponds to a group of the storage drives such that a failure corresponding to the failure domain causes failure of each storage drive of the group, the method further comprising:

for each failure sub-domain type of a plurality of failure sub-domain types, grouping the storage drives by failure sub-domains corresponding to the failure sub-domain type;

determining a number of required failure domains from the request;

combining the failure sub-domains into which the storage drives have been grouped until the storage drives have been grouped into a number of the failure domains equal to the number of required failure domains, taking into account redundancy importance across the failure domains.

15. The non-transitory computer-readable data storage medium of claim 13, wherein the method further comprises:

filtering the storage drives by the native capabilities thereof to remove from selection consideration any storage drive having insufficient native capabilities,

and wherein selecting the set of storage drives comprises selecting the storage drives of the set to maximize sameness in the native capabilities thereof.

16. The non-transitory computer-readable data storage medium of claim 13, wherein the method further comprises:

filtering the storage drives by the available capacities thereof to remove from selection consideration any storage drive having insufficient available capacity,

and wherein selecting the set of storage drives comprises selecting the storage drives of the set to one of minimize or maximize remaining capacity on each storage drive after the drive volume has been provisioned on the storage drive.

17. The non-transitory computer-readable data storage medium of claim 13, wherein the method further comprises, prior to selecting the set of drives, one of:

retrieving from each storage drive dynamic characteristics maintained by the storage drive itself that are indicative of the dynamic performance of the storage drive;

measuring the dynamic performance between the SAN manager and each storage drive;

measuring the dynamic performance between the client computing device and each storage drive.

18. A storage-area network (SAN) system comprising:

a plurality of fabric-attachable storage drives directly connected to a fabric, each storage drive of a set of the storage drives to provision and operate a drive volume;

a client computing device directly connected to the fabric, and that is to execute a SAN software agent to create, mount, and use a logical volume realized by drive volumes of the set of storage drives; and

a SAN manager directly connected to the fabric and that is to manage the drive volumes of the storage drives, to manage the logical volume that the SAN software agent operates, and is to select the set of the storage drives from the storage drives directly connected to the fabric on which the SAN software agent is to create, mount, and use the logical volume on drive volumes provisioned thereon.

19. The SAN system of claim 18, wherein the SAN manager is to select the set of the storage drives based on native capabilities, available capacities, and dynamic performance thereof, and taking into account failure domains in which the storage drives are located.