Efficient mechanism to replicate data for multiple controllers

• US 9,600,378 B2
• Filed: 01/29/2015
• Issued: 03/21/2017
• Est. Priority Date: 01/29/2015
• Status: Active Grant

• Current Assignee: Dell Products LP (Dell Technologies Inc.)
• Original Assignee: Dell Products LP (Dell Technologies Inc.)
• Inventors: Nelogal, Chandrashekar, Giannoules, James P.
• Primary Examiner(s): DOAN, DUC T
• Application Number:
  US14/608,821
• Publication Number:
  US 20160224437A1
• Time in Patent Office: 782 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06F 11/1435 using file system or storage system metadata
• CPC Class Codes: G06F 11/1451 by selection of backup contents
• CPC Class Codes: G06F 11/1469 Backup restoration techniques
• CPC Class Codes: G06F 11/1658 Data re-synchronization of a redundant component, or initial sync of replacement, additional or spare unit
• CPC Class Codes: G06F 11/2089 Redundant storage control functionality
• CPC Class Codes: G06F 2201/84 Using snapshots, i.e. a logical point-in-time copy of the data
• CPC Class Codes: G06F 2201/845 Systems in which the redundancy can be transformed in increased performance

Abstract

An information handling system and method allows implementation of fault-tolerant storage subsystems using multiple storage controllers not themselves originally designed to support the redundancy of such fault-tolerant storage subsystems. In accordance with one embodiment, uncommitted data is efficiently and rapidly replicated across multiple commodity storage controllers, enabling faster and less expensive fault-tolerant storage subsystems. A redundant storage controller system can improve the efficiency of data replication while providing failure protection against controller failure. A redundant storage controller system using shared memory commonly accessible to the storage controllers can be enhanced to replicate data within host memory regions to protect against non-volatile memory failure. In accordance with at least one embodiment, an efficient data replication mechanism can be provided between storage controllers using off-the-shelf hardware.

Claims

1. A method comprising:

issuing, at a host processor, a write command to a primary storage controller;

receiving, at the primary storage controller, the write command;

transferring, at the primary storage controller, write data associated with the write command to a primary storage controller memory local to the primary storage controller;

resending, at the host processor, the write command to a peer storage controller, wherein the peer storage controller provides redundancy for the primary storage controller;

receiving, at the peer storage controller, the write command;

transferring, at the peer storage controller, the write data associated with the write command to a peer storage controller memory local to the peer storage controller;

writing, by the primary storage controller, the write data from the primary storage controller memory to a storage subsystem; and

instructing, by the primary storage controller, the peer storage controller to discard the write data stored in the peer storage controller memory.

2. The method of claim 1 further comprising:

discarding, by the peer storage controller, the write data stored in the peer storage controller memory.

3. The method of claim 1 further comprising:

notifying, by the primary storage controller, the host processor of completion of the write command after the transferring, at the primary storage controller, the write data associated with the write command to the primary storage controller memory.

4. The method of claim 1 further comprising:

acknowledging, by the peer storage controller, receipt of the write command to the host processor.

5. The method of claim 1 wherein the writing, by the primary storage controller, the write data from the primary storage controller memory to the storage subsystem further comprises:

generating parity data.

6. The method of claim 1 wherein the issuing, at the host processor, the write command to the primary storage controller comprises issuing a scatter-gather list (SGL), and wherein the resending, at the host processor, the write command to the peer storage controller comprises:

translating, at the host processor, the SGL into a translated SGL for the peer storage controller; and

sending, by the host processor, the translated SGL to the peer storage controller.

7. The method of claim 1 further comprising:

when the primary storage controller fails, establishing the peer storage controller as a newly established primary storage controller and writing, by the newly established primary storage controller, the write data from the newly established primary storage controller memory of the newly established primary storage controller to the storage subsystem.

8. An information handling system comprising:

a host processor including a host memory for storing instructions executable by the host processor to cause the host processor to issue a write command to a primary storage controller and to resend the write command to a peer storage controller; and

the primary storage controller including a primary storage controller memory for storing instructions executable by the primary storage controller to cause the primary storage controller to receive the write command, to transfer write data associated with the write command to the primary storage controller memory, and to write the write data from the primary storage controller memory to a storage subsystem,

wherein the peer storage controller provides redundancy for the primary storage controller, the peer storage controller including peer storage controller memory for storing instructions executable by the peer storage controller to cause the peer storage controller to receive the write command, and to transfer the write data associated with the write command to the peer storage controller memory, wherein the instructions executable by the primary storage controller further comprise instructions to cause the primary storage controller to instruct the peer storage controller to discard the write data stored in the peer storage controller memory.

9. The information handling system of claim 8 wherein the instructions executable by the peer storage controller further comprise instructions to cause the peer storage controller to discard the write data stored in the peer storage controller memory.

10. The information handling system of claim 8 wherein the instructions executable by the primary storage controller further comprise instructions to cause the primary storage controller to notify the host processor of completion of the write command after the primary storage controller transfers the write data associated with the write command to the primary storage controller memory.

11. The information handling system of claim 8 wherein the instructions executable by the peer storage controller further comprise instructions to cause the peer storage controller to acknowledge receipt of the write command to the host processor.

12. The information handling system of claim 8 wherein the instructions executable by the primary storage controller to cause the primary storage controller to write the write data from the primary storage controller memory to a storage subsystem further comprise instructions executable by the primary storage controller to cause the primary storage controller to generate parity data.

13. The information handling system of claim 8 wherein the instructions executable by the host processor to cause the host processor to issue the write command to the primary storage controller further comprise instructions to cause the host processor to issue a scatter-gather list (SGL), and wherein the instructions executable by the host processor to cause the host processor to resend the write command to the peer storage controller further comprise instructions executable by the host processor to translate the SGL into a translated SGL for the peer storage controller and to send the translated SGL to the peer storage controller.

14. The information handling system of claim 8 wherein the instructions executable by the peer storage controller further comprise instructions executable by the peer storage controller, when the primary storage controller fails, to establish the peer storage controller as a newly established primary storage controller and to write, by the newly established primary storage controller, the write data from the newly established primary storage controller memory of the newly established primary storage controller to the storage subsystem.

15. A method comprising:

issuing, at a host processor, a write command to a primary storage controller;

receiving, at the primary storage controller, the write command;

transferring, by the primary storage controller, write data associated with the write command to a host memory region of a host memory local to the host processor, wherein the host memory region is commonly accessible to the primary storage controller and to a peer storage controller, wherein the peer storage controller provides redundancy for the primary storage controller;

informing, by the primary storage controller, a peer storage controller of the write data stored in the host memory region;

generating, by the primary storage controller, parity data in response to the write command;

storing, by the primary storage controller, the parity data in the host memory region;

updating, by the primary storage controller, metadata for use by the peer storage controller;

writing, by the primary storage controller, the write data from the host memory region to a storage subsystem; and

updating, by the primary storage controller, the metadata in response to the writing.

16. The method of claim 15 wherein the writing further comprises:

generating journal entries; and

storing the journal entries in the host memory region.

17. The method of claim 15 further comprising:

transferring, at the primary storage controller, write data associated with the write command to a primary storage controller memory local to the primary storage controller.

18. The method of claim 17 further comprising:

upon failure of the primary storage controller, writing, by a peer storage controller, the write data from the host memory region to the storage subsystem; and

upon failure of the host memory, writing, by the primary storage controller, the write data from the primary storage controller memory to the storage subsystem.

19. The method of claim 18 further comprising:

instructing, by the primary storage controller, the peer storage controller to discard the write data stored in the peer storage controller memory.

20. The method of claim 15 further comprising:

when the primary storage controller fails, establishing the peer storage controller as a newly established primary storage controller and writing, by the newly established primary storage controller, the write data from the host memory region to the storage subsystem.