STORAGE DEVICE DRIVERS AND CLUSTER PARTICIPATION

• US 20130067188A1
• Filed: 09/12/2011
• Published: 03/14/2013
• Est. Priority Date: 09/12/2011
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Mehra, Karan, Nagar, Rajeev, Rajpal, Shiv, Moss, Darren, D'Amato, Andrea, Warwick, Alan, Kuznetsov, Vyacheslav
• Granted Patent: US 8,886,910 B2
• Time in Patent Office: Days
• US Class Current: 711/170
• CPC Class Codes: G06F 3/0605 by facilitating the interaction with a user or administrator
• CPC Class Codes: G06F 3/0632 by initialisation or re-initialisation of storage systems
• CPC Class Codes: G06F 3/0689 Disk arrays, e.g. RAID, JBOD
• CPC Class Codes: G06F 9/4411 Configuring for operating with peripheral devices; Loading of device drivers

Abstract

The representation of storage devices on computers (e.g., as logical volumes) may be complicated by the pooling of multiple storage devices in order to apply redundancy plans such as mirroring and checksumming. Presented herein is a storage device driver configured to operate as a storage device interface generating representations of the storage regions of the storage devices; to claim those regions as a storage controller; and to expose pooled storage regions as logical disks. Additionally, the storage device driver may support the inclusion of storage devices in a cluster, comprising nodes that may be appointed as managers of the storage pool configuration; as managers of the storage devices; as owners having exclusive read/write access to the storage pool or cluster resources; and as cluster resource writers having excusive write access to a cluster resource. The nodes of the cluster may interoperate to share the storage devices while avoiding write conflicts.

Claims

1. A storage device driver configured to represent storage devices accessible to a computer, the storage device driver comprising:

a storage device interface component configured to generate a storage region representation of respective storage regions of storage devices;

a storage controller driver component configured to, for respective storage pools comprising at least one storage region of at least one storage device;

claim the storage region representations of the storage regions of the storage devices involved in the storage pool;

apply a storage device relationship to the storage pool; and

generate a pooled storage region representation of the storage pool; and

a storage port driver component configured to, for respective pooled storage region representations, generate a logical representation of the pooled storage region representation.

2. The storage device driver of claim 1:

respective storage pools represented on respective storage devices as a storage pool configuration;

the computer included in a cluster comprising nodes having access to the storage devices comprising a storage pool;

for respective storage pools, a storage pool manager having been selected among the nodes of the cluster; and

the storage controller driver component configured to;

upon receiving a request to access a storage device, access the storage device through the storage device interface component; and

upon receiving a request to update a storage pool configuration for a storage device;

determine whether the node is the storage pool manager of the storage pool including the node;

upon determining that the node is not the storage pool manager, forward the request to the storage pool manager; and

upon determining that the node is the storage pool manager, update the storage pool configuration through the storage device interface component for the storage device.

3. A method of representing storage devices accessible to a computer having a processor, the method comprising:

install on the computer a storage device driver comprising instructions configured to, when executed on the processor, cause the computer to;

generate a storage region representation of respective storage regions of storage devices;

for respective storage pools comprising at least one storage region of at least one storage device;

claim the storage region representations of the storage regions of the storage devices involved in the storage pool;

apply a storage device relationship to the storage pool; and

generate a pooled storage region representation of the storage pool; and

for respective pooled storage region representations, generate a logical representation of the storage region representation.

4. The method of claim 3:

the computer configured to present a storage device enumeration of storage devices accessible to the computer;

claiming the storage device representations of storage devices involved in a storage pool comprising;

removing the storage device from the storage device enumeration; and

generating the storage region representation of a storage pool comprising;

adding the pooled storage region representation to the storage device enumeration.

5. The method of claim 3:

the computer included in a cluster comprising nodes having access to the storage devices comprising a storage pool; and

the instructions configured to, for respective storage pools;

among the nodes of the cluster, select a storage pool owner having exclusive access to the storage pool;

upon detecting a substitution event involving the storage pool owner, among available nodes of the cluster, select a substitute storage pool owner;

upon the node being selected as the storage pool owner, indicate the storage pool as accessible to the node;

upon another node being selected as the storage pool owner, indicate the storage pool as inaccessible to the node.

6. The method of claim 5, selecting the storage pool owner comprising:

for respective storage pools having a redundancy plan, selecting a storage pool owner having exclusive access to the storage pool.

7. The method of claim 5, the substitution event selected from a substitution event set comprising:

a load rebalancing event indicating a rebalancing from a first node of the cluster to a second node of the cluster;

a disconnection event indicating a disconnection of a node from the cluster; and

a failure event indicating a failure of a node.

8. The method of claim 3:

respective storage pools represented on respective storage devices as a storage pool configuration;

the computer included in a cluster comprising nodes having access to the storage devices comprising a storage pool; and

the instructions configured to, for respective storage pools;

among the nodes of the cluster, select a storage pool manager configured to update the storage pool configuration; and

upon detecting a substitution event involving the storage pool manager, among available nodes of the cluster, select a substitute storage pool manager.

9. The method of claim 8, the substitution event selected from a substitution event set comprising:

a load rebalancing event indicating a rebalancing from a first node of the cluster to a second node of the cluster;

a disconnection event indicating a disconnection of a node from the cluster; and

a failure event indicating a failure of a node.

10. The method of claim 8, the instructions configured to, upon receiving a request to update a storage pool configuration for a storage device:

determine whether the node is the storage pool manager of the storage pool including the node;

upon determining that the node is not the storage pool manager, forward the request to the storage pool manager; and

upon determining that the node is the storage pool manager, update the storage pool configuration on the storage device.

11. The method of claim 10:

the cluster comprising a broadcast interface to broadcast updates to the cluster; and

the instructions configured to, upon updating a storage pool configuration on a storage device, invoke the broadcast interface to notify other nodes of the cluster of the update to the storage pool configuration.

12. The method of claim 8, the instructions configured to, upon receiving a notification of an update to the storage pool configuration, read the storage pool configuration.

13. The method of claim 8:

the storage pool configuration of respective storage pools comprising a clustering participation indicator; and

the instructions configured to;

upon a storage pool joining a cluster, set the clustering participation indicator of the storage pool configuration of the storage pool;

upon a storage pool leaving a cluster, clear the clustering participation indicator of the storage pool configuration of the storage pool; and

upon receiving a request to update the storage pool configuration of a storage pool, if the clustering participation indicator of the storage pool is set and the computer is not the storage pool manager of the storage pool, prohibit updating the storage pool configuration of the storage pool.

14. The method of claim 8:

respective storage devices of the storage pool comprising a registration list comprising at least zero registrations specifying a node that is authorized to write to the storage regions of the storage pool on the storage device; and

the instructions configured to, before writing to a storage device storing at least one storage region of the storage pool, add the node to the registration list of the storage device.

15. The method of claim 14, the instructions configured to:

for respective storage devices storing at least one storage region, among the nodes of the cluster, select a storage device manager configured to manage the registration list of the storage device; and

upon determining that the computer is the storage device manager of a storage device;

add a reservation to the registration list identifying the node of the cluster selected as the storage device manager of the storage device;

monitor the registration list of the storage device; and

upon detecting in the registration list a registration of a computer that is not a node of the cluster, remove the registration of the computer from the registration list.

16. The method of claim 15, the storage device manager comprising the storage pool manager of the storage pool.

17. The method of claim 14, the instructions configured to, upon receiving a request to access a storage region:

determine whether the storage region comprises a redundancy plan; and

upon determining that the storage region comprises a redundancy plan;

read the registration lists of the storage devices;

for respective other nodes having a registration on the registration list, request the other node to unmount the storage region; and

upon receiving from the other nodes an unmount confirmation, access the storage region of the storage device; and

upon receiving from a node a request to unmount a storage region;

if the storage region is mounted, unmount the storage region; and

send an unmount confirmation to the node.

18. The method of claim 8:

at least one storage pool comprising at least one cluster resource; and

the instructions configured to;

for respective cluster resources, among the nodes of the cluster, select a cluster resource owner having exclusive read and write access to the cluster resource; and

upon detecting a substitution event involving a node of the cluster;

identify the cluster resources for which the node comprised the cluster resource owner; and

among available nodes of the cluster, for respective cluster resources, select a substitute cluster resource owner.

19. The method of claim 8:

at least one storage pool comprising at least one cluster resource; and

the instructions configured to;

for respective cluster resources, among the nodes of the cluster, select a cluster resource writer having exclusive write access to the cluster resource; and

upon detecting a substitution event involving a node of the cluster;

identify the cluster resources for which the node comprised the cluster resource writer; and

among available nodes of the cluster, for respective cluster resources, select a substitute cluster resource writer.

20. The method of claim 19, the cluster resources selected from a cluster resource set comprising:

at least one logical representation manifested from the storage pool, and

the storage pool configuration for the storage pool.

21. The method of claim 19:

respective storage regions of respective storage devices having a maintenance region configured to store health information about the storage region; and

the instructions configured to, upon detecting a health update of a storage region, update the maintenance region to record the health update.

22. The method of claim 21, the instructions configured to, if the computer is the storage pool manager of the storage pool:

read the maintenance region to identify storage regions to be reallocated in response to a health update;

reallocate the storage regions; and

update the storage pool configuration.

23. The method of claim 22, the instructions configured to, upon detecting an update of the storage pool configuration comprising an allocation of at least one reallocated storage region, if the computer is the cluster resource writer of the storage region, initiate a repair operation to the reallocated storage regions.

24. A computer-readable storage medium comprising instructions that, when executed by a processor of a device having access to storage devices, represent the storage devices by:

installing on the computer a storage device driver comprising instructions configured to, when executed on the processor, cause the computer to;

generating a storage region representation of respective storage regions of storage devices;

for respective storage pools comprising at least one storage region of at least one storage device;

claiming the storage region representations of the storage regions of the storage devices involved in the storage pool;

applying the storage device relationship to the storage pool; and

generating a pooled storage region representation of the storage pool; and

for respective pooled storage region representations, generating a logical representation of the storage region represented by the storage region representation.