Systems, apparatus, and methods for processing I/O requests

• US 20060047903A1
• Filed: 08/24/2004
• Published: 03/02/2006
• Est. Priority Date: 08/24/2004
• Status: Active Grant

• Current Assignee: Veritas Technologies, LLC (Whitehouse Group Ltd.)
• Original Assignee: Symantec Operating Corporation (NortonLifeLock Inc.)
• Inventors: Passerini, Ron
• Granted Patent: US 7,827,362 B2
• Time in Patent Office: Days
• US Class Current: 711/114
• CPC Class Codes: G06F 3/0622 in relation to access
• CPC Class Codes: G06F 3/0656 Data buffering arrangements
• CPC Class Codes: G06F 3/0659 Command handling arrangements, e.g. command buffers, queues, command scheduling
• CPC Class Codes: G06F 3/0683 Plurality of storage devices

Abstract

A method, apparatus, and system for accessing units of storage in at least one logical unit by processing I/O requests directed to the logical units using a LUN queue and an operation-type queue. By using the queues to process the I/O requests, the requests can be processed without address collisions.

Claims

1. A system for processing I/O requests directed to at least one logical unit of storage, the system comprising:

an operation memory for storing a plurality of ordered sets of operations, each set associated with an I/O request;

a processor, in communication with the operation memory, for queuing the operations stored in the operation memory;

first queues, in communication with the processor, for queuing the operations based on an identification of a target logical unit; and

second queues, in communication with the processor, for queuing operations based on an operation type.

2. The system of claim 1, wherein a first operation in the ordered set of operations is processed in one of the first queues.

3. The system of claim 2 wherein operations subsequent to the first operation in the ordered set are processed in one of the second queues.

4. The system of claim 1 wherein the operations in each of the first queues are processed in chronological order based on a time of receipt of the I/O requests associated with the operations in the first queues.

5. The system of claim 1 wherein the operations in each of the second queues are batch processed.

6. A storage management device for accessing at least one logical unit, comprising:

a plurality of I/O processing modules, comprising;

a host interface in communication with a host for receiving I/O requests, and extracting a set of ordered operations associated with each I/O request;

an I/O manager in communication with the host interfaces of the processing modules, the I/O manager for processing the extracted operations first using a LUN queue and thereafter using an operation-type queue; and

a physical store interface in communication with the I/O managers of the processing modules, the physical store interface for communicating with at least one logical unit in response to the processed operations; and

an internal network for communication among the processing modules.

7. The storage management device of claim 6 wherein the host interface classifies the I/O requests based on a logical unit identification associated with the I/O request.

8. The storage management device of claim 7 wherein operations directed to a first logical unit are stored in a first LUN queue and operations directed to a second logical unit are stored in a second LUN queue.

9. The storage management device of claim 7 wherein write operations directed to the logical units are indexed with an indexed association.

10. The storage management device of claim 9 wherein the indexed association comprises:

a data store ID;

a data store address;

a length associated with the data store address;

a logical unit ID;

a logical unit address;

a length associated with the logical unit address; and

a time.

11. The storage management device of claim 6 wherein operations in each LUN queue are organized in a data structure that associates operations stored in the LUN queue with one another when the operations are directed to a common range of logical block addresses in the same data store.

12. A method of accessing units of storage in a logical unit, the method comprising the steps of:

receiving I/O requests from a host;

classifying received I/O requests based on a target logical unit associated with the I/O requests;

extracting ordered sets of operations associated with the I/O requests;

processing the extracted operations using a LUN queue and an operation-type queue; and

communicating with at least one logical unit to access units of storage in response to the I/O requests.

13. The system of claim 12, wherein a first operation in the ordered set of operations is processed in at least one of the LUN queues.

14. The system of claim 13 wherein operations subsequent to the first operation in the ordered set are processed in at least one of the operation-type queues.

15. The system of claim 13 wherein a plurality of first operations in a LUN queue are processed in an order in which the I/O requests were received.

16. The system of claim 14 wherein the subsequent operations are organized into one or more batches for processing.

17. A storage network, comprising:

a plurality of logical units; and

a storage management device, comprising;

a plurality of I/O processing modules, comprising;

a host interface in communication with a host for receiving I/O requests and extracting operations associated with each I/O request;

an I/O manager in communication with the host interfaces of the processing modules, the I/O manager for processing the extracted operations using a LUN queue and an operation-type queue; and

a physical store interface in communication with the I/O managers of the processing modules, the physical store interface for communicating with at least one logical unit in response to the processed operations; and

an internal network for communication among the processing modules.

18. A method of executing I/O requests by a storage management device, the method comprising the steps of:

receiving an I/O request directed to a first logical unit;

specifying an ordered set of operations in response to the received I/O request;

sending a first operation from the ordered set of operations to a LUN queue associated with the first logical unit;

processing the first operation; and

sending a subsequent operation from the ordered set of operations to a second queue associated with a type of the subsequent operation.

19. The method of claim 18 wherein the subsequent operation is a read before-image operation.

20. The method of claim 19, further comprising the steps of:

receiving additional I/O requests directed to the first logical unit;

specifying an additional ordered set of operations in response to each additional I/O request;

sending a first operation from each additional ordered set of operations to the LUN queue associated with the first logical unit; and

processing the first operation from each additional ordered set of operations in the order received by the LUN queue associated with the first logical unit.

21. The method of claim 18, further comprising the steps of:

receiving an additional I/O request directed to an additional logical unit; and

sending the first operation from the additional I/O request to a LUN queue associated with the additional logical unit.

22. The method of claim 21 wherein the processing step, further comprises:

determining a processing order for operations in a plurality of LUN queues based on an amount of time that operations spend in each of the plurality of LUN queues.

23. The method of claim 22, further comprising the step of:

employing a fairness algorithm to determine the processing order.