Network Based Virtualization Performance

• US 20100023724A1
• Filed: 10/05/2009
• Published: 01/28/2010
• Est. Priority Date: 04/26/2005
• Status: Active Grant

• Current Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Original Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Inventors: Kumar, Sanjaya, Sharma, Samar, Bhandari, Rajesh
• Granted Patent: US 8,644,174 B2
• Time in Patent Office: Days
• US Class Current: 711/171
• 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

The disclosed embodiments support improvements in network performance in networks such as storage area networks. This is particularly important in networks such as those implementing virtualization. These improvements, therefore, support improved mechanisms for performing processing in network devices such as switches, routers, or hosts. These improvements include various different mechanisms which may be used separately or in combination with one another. These mechanisms include methods and apparatus for processing traffic in an arbitrated loop, performing striping to support fairness and/or loop tenancy, performing configuration of network devices such as switches to enable virtualization to be performed closest to the storage device (e.g., disk), ascertaining a CPU efficiency that quantifies the impact of virtualization on a processor, and configuring or accessing a striped volume to account for metadata stored in each storage partition.

Claims

1. A method, comprising:

identifying a stripe size, the stripe size being a size of data to be written to or read from one of a plurality of columns;

determining a host IO size, the host IO size being a size of data to be received from or transmitted to a host, thereby enabling the data to be stored to or read from storage including a plurality of columns;

setting the host IO size to identify an amount of data to be received from or transmitted to the host;

obtaining data from the host or transmitting data to the host, wherein the size of the data obtained or transmitted conforms to the host IO size; and

performing striping across the plurality of columns for the data obtained from the host or transmitted to the host such that data written to or read from one of the plurality of columns conforms to the stripe size.

2. The method as recited in claim 1, further comprising:

receiving a command from the host, the command being a read or write command.

3. The method as recited in claim 1, further comprising:

sending a XFR_RDY command identifying the host IO size to the host, wherein the method is performed in a Fibre Channel network.

4. The method as recited in claim 3, further comprising:

receiving data from the host, wherein the size of the data is equal to the host IO size.

5. The method as recited in claim 1, wherein each of the plurality of columns is a disk.

6. The method as recited in claim 1, wherein determining the host IO size comprises:

identifying a minimum of a size of the remaining data associated with the host IO and a product of the stripe size and the number of columns;

when the minimum is a size of the remaining data associated with the host IO, the host IO size is the size of the remaining data; and

when the minimum is a product of the stripe size and the number of columns, the host IO size is the product of the stripe size and the number of columns.

7. The method as recited in claim 6, wherein determining, setting the host IO size and performing striping are performed dynamically until striping has been completed for the data.

8. The method as recited in claim 1, wherein determining the host IO size comprises:

calculating the host IO size using the number of columns, the stripe size and the remaining data associated with the host IO.

9. The method as recited in claim 1, wherein the method is performed in a network implementing virtualization.

10. A method, comprising:

receiving a command from a host, the command being a request to store or retrieve data, the command indicating a size of the data to be stored or retrieved;

ascertaining whether the size of the data associated with the command is greater than a product of a configured stripe size and number of a plurality of columns; and

reading data from the plurality of columns or writing data to the plurality of columns, wherein reading data or writing data is performed in accordance with the configured stripe size and the command.

12. The method as recited in claim 10, further comprising:

when the size of the data associated with the command is greater than the product of the configured stripe size and the number of columns, determining a host IO size, setting a host IO size, and obtaining data conforming to the host IO size from the host or transmitting data conforming to the host IO size to the host by performing striping across a plurality of columns, wherein the data written to or read from each of the plurality of columns conforms to the configured stripe size.

13. The method as recited in claim 10, wherein the command is a read command.

14. The method as recited in claim 12, further comprising:

sending a request for data to the host, wherein the size of data requested conforms to the host IO size.

15. The method as recited in claim 10, wherein receiving the command from the host comprises:

receiving a request from the host that identifies a virtual storage location and wherein the plurality of columns are associated with a plurality of physical storage locations.

16. The method as recited in claim 12, wherein determining, setting, and obtaining or transmitting are performed dynamically until striping has been completed for the data.

17. The method as recited in claim 16, wherein determining the host IO size comprises:

identifying a minimum of a size of the remaining data associated with the command and a product of the stripe size and the number of columns;

when the minimum is a size of the remaining data associated with the command, the host IO size is the size of the remaining data; and

when the minimum is a product of the stripe size and the number of columns, the host IO size is a product of the stripe size and the number of columns.

18. The method as recited in claim 10, wherein the command is a write command.

11. The method as recited in 10, further comprising:

when the size of the data associated with the command is not greater than the product of the configured stripe size and the number of columns, performing striping according to the configured stripe size.

19. A computer-readable storage medium storing thereon computer-readable instructions, comprising:

instructions for identifying a striped volume in which both metadata and data are stored in a storage area network;

instructions for identifying a size of the metadata stored for each partition in the striped volume;

instructions for identifying a stripe size for data stored in the storage area network; and

instructions for calculating a start address for a partition in the striped volume, wherein the start address for the partition in the striped volume is a stripe aligned address at which data is stored less a sum of the metadata region size for the partition.

20. An apparatus, comprising:

a processor; and

a memory, at least one of the processor or the memory being adapted for;

identifying the striped volume in which both metadata and data are stored in a storage area network;

identifying a size of the metadata stored for each partition in the striped volume;

identifying a stripe size for data stored in the storage area network; and

calculating an application data start address identifying a start address at which data is stored, wherein the application data start address is the sum of the start address for the partition in the striped volume and the metadata region size for the partition.