Shared storage I/O elimination through mapping client integration into a hypervisor

• US 8,407,448 B1
• Filed: 05/06/2008
• Issued: 03/26/2013
• Est. Priority Date: 05/06/2008
• Status: Active Grant

• Current Assignee: Emc IP Holding Company LLC (Dell Technologies Inc.)
• Original Assignee: EMC Corporation (Dell Technologies Inc.)
• Inventors: Hayden, John, Jiang, Xiaoye
• Primary Examiner(s): Schnee, Hal
• Application Number:
  US12/115,617
• Time in Patent Office: 1,785 Days
• Field of Search: 711/112, 711/162, 711/202, 711/203, 711/E12.019
• US Class Current: 711/203
• CPC Class Codes: G06F 12/1045 associated with a data cache
• CPC Class Codes: G06F 12/109 for multiple virtual address spaces, e.g. segmentation G06F12/1036 takes precedence
• CPC Class Codes: G06F 2009/45579 I/O management, e.g. providing access to device drivers or storage
• CPC Class Codes: G06F 9/45533 Hypervisors; Virtual machine monitors

Abstract

This invention is a system and a method for performing an I/O in a virtual data storage environment using a new architecture. The system of performing an I/O includes a mapping client integrated into a client of the storage server which in communication with the mapping server included in the storage server retrieves the mapping of the special data sharing storage objects and caches the shared objects in the data cache include in the client environment. The method of accessing the data sharing storage objects by one or more applications running on a client reduces the number of I/O on the storage objects by caching the storage objects in the data cache and bringing the knowledge of data sharing into the client environment.

Claims

1. A system for performing an I/O in a storage virtualization environment, the system comprising:

a storage server in the storage virtualization environment, including a mapping service, wherein the storage server in communication with the mapping service organizes one or more storage objects as a set of related objects indicating a portion of data that is shared among the one or more storage objects;

a data storage system in the storage virtualization environment, wherein the data storage system in communication with the storage server provides a pool of storage resources to the storage server for storing the one or more storage objects as a set of virtual devices that share from the pool of storage resources; and

a client of the storage server including a mapping client for the mapping service and a data cache, wherein the client of the storage server in communication with the storage server and the data storage system presents a storage object of the one or more storage objects as a logical addressable storage object to one or more applications running on the client of the storage server and the mapping client provides the one or more application a mapping between a logical addressable storage object to its physical location on the data storage system and uses the data cache to store the physical location and contents of the one or more storage objects;

wherein the one or more storage objects is represented by one or more files;

wherein the storage server is represented by a file server in communication with the data storage system;

wherein the data storage system provides the logical disk storage to the file server for storing the one or more files;

wherein the one or more files are organized as a version set by the file server indicating a set of physical blocks that are shared among the one or more files;

wherein the one or more files organized as the version set represents the logical disk storage for respective virtual machines of a group of one or more virtual machines in the storage virtualization environment; and

wherein the version set is a sparse snapshot of the file of the one or more files.

2. The system of claim 1, wherein the mapping service is represented by a mapping server included in the file server, wherein the mapping server contains the mapping between a logical address of a block and a physical address of the block belonging to a file of the one or more files organized as a version set on the file server.

3. The system of claim 1, wherein the data cache is represented by a physical block cache, wherein the physical block cache stores address and content of a physical block of a data stored on the data storage system.

4. The system of claim 1, wherein the application running on the client of the storage server is represented by a virtual machine, wherein the virtual machine is a representation of a real machine using software that provides an operating system environment.

5. The system of claim 1, wherein the client of the storage server is represented by a Hypervisor, wherein the Hypervisor in communication with the file server and the data storage system presents the one or more files organized as a version set on the file server as volumes with a Logical Unit Number to one or more virtual machines running on the Hypervisor and the mapping client provides the virtual machine of the one or more virtual machines a mapping between a logical address of a block and the physical address of the block belonging to a file of the one or more files and uses the physical block cache to store the address and content of the block stored on the data storage system.

6. The system of claim 5, wherein the mapping client included in the Hypervisor further includes the information if a physical block is shared by one or more logical blocks.

7. The system of claim 5, wherein the file of one or more files organized as a version set on the file server represents a boot and a data volume for one or more virtual machines running on the Hypervisor.

8. A method for performing an I/O on a storage object of one or more storage objects in a storage virtualization environment, the environment including a storage server, a data storage system and a client of a storage server, wherein the data storage system in communication with the storage server provides a pool of storage resources to the storage server for storing the one or more storage objects as a set of virtual devices that share from the pool of storage resources organized as a set of related objects indicating a portion of data that is shared among the one or more storage objects, wherein the storage object of the one or more storage object is presented by the client of the storage server as a logical addressable object to one or more applications running on the client, the method comprising the steps of:

retrieving from a mapping client a physical address for a logical addressable storage object on which I/O is generated by an application, wherein the mapping client resides in the client of the storage server and provides the application a mapping of the logical addressable storage object to its physical location on the data storage system;

examining a data cache included in the client of the storage server to determine the availability of the storage object in the data cache, wherein the data cache contains the physical address and content of the one or more storage objects accessed by previous I/O operations; and

completing the I/O on finding the storage object in the data cache;

wherein the one or more storage objects is represented by one or more files;

wherein the storage server is represented by a file server in communication with the data storage system;

wherein the data storage system provides the logical disk storage to the file server for storing the one or more files;

wherein the one or more files are organized as a version set by the file server indicating a set of physical blocks that are shared among the one or more files;

wherein the one or more files organized as the version set represents the logical disk storage for respective virtual machines of a group of one or more virtual machines in the storage virtualization environment; and

wherein the version set is a sparse snapshot of the file of the one or more files.

9. The method of claim 8, wherein providing, by the mapping client, the application a mapping of the logical addressable storage object to its physical location on the data storage system comprises the steps of:

sending the I/O request to a mapping service on not finding the physical address for the logically addressable storage object in the mapping client, wherein the mapping service is included in the storage server and translates the logical address of a storage object to a physical address on the data storage system and returns the address to the mapping client;

sending the I/O request to the data storage system to retrieve the storage object from the physical storage; and

caching the physical address and the content of the storage object block in a data cache included in the client of the storage server.

10. The method of claim 8, wherein the step of caching the storage object in the data cache is performed in response to a determination of an attribute of a storage object indicating that the portion of the data is shared among one or more storage objects.

11. The method of claim 8, wherein the mapping service is represented by a mapping server included in the file server, wherein the mapping server contains the mapping between a logical address of a block and a physical address of the block belonging to a file of the one or more files organized as a version set on the file server.

12. The method of claim 8, wherein the data cache is represented by a physical block cache, wherein the physical block cache stores address and content of a physical block of a data stored on the data storage system.

13. The method of claim 8, wherein the application running on the client of the storage server is represented by a virtual machine, wherein the virtual machine is a representation of a real machine using software that provides an operating system environment.

14. The method of claim 8, wherein the client of the storage server is represented by a Hypervisor, wherein the Hypervisor in communication with the file server and the data storage system presents the one or more files organized as a version set on the file server as volumes with a Logical Unit Number to one or more virtual machines running on the Hypervisor and the mapping client provides the virtual machine of the one or more virtual machines a mapping between a logical address of a block and the physical address of the block belonging to a file of the one or more files and uses the physical block cache to store the address and content of the block stored on the data storage system.

15. The method of claim 14, wherein the mapping client included in the Hypervisor further includes the information if a physical block is shared by one or more logical blocks.

16. The method of claim 8, wherein the file of the one or more files organized as a version set on the file server represents a boot and a data volume for one or more virtual machines running on the Hypervisor.

17. A system for performing an I/O in a storage virtualization environment, the system comprising:

a storage server in the storage virtualization environment, including a mapping service, wherein the storage server in communication with the mapping service organizes one or more storage objects as a set of related objects indicating a portion of data that is shared among the one or more storage objects;

a data storage system in the storage virtualization environment, wherein the data storage system in communication with the storage server provides a pool of storage resources to the storage server for storing the one or more storage objects as a set of virtual devices that share from the pool of storage resources;

a client of the storage server including a mapping client for the mapping service and a data cache, wherein the client of the storage server in communication with the storage server and the data storage system presents a storage object of the one or more storage objects as a logical addressable storage object to one or more applications running on the client of the storage server and the mapping client provides the one or more application a mapping between a logical addressable storage object to its physical location on the data storage system and uses the data cache to store the physical location and contents of the one or more storage objects; and

a program logic in communication with the data storage system and the file server for carrying out the steps of;

retrieving from the mapping client a physical address for a logical addressable storage object on which I/O is generated by an application, wherein the mapping client resides in the client of the storage server and provides the application a mapping of the logical addressable storage object to its physical location on the data storage system;

examining the data cache included in the client of the storage server to determine the availability of the storage object in the data cache, wherein the data cache contains the physical address and content of the one or more storage objects accessed by previous I/O operations; and

completing the I/O on finding the storage object in the data cache wherein the one or more storage objects is represented by one or more files;

wherein the storage server is represented by a file server in communication with the data storage system;

wherein the data storage system provides the logical disk storage to the file server for storing the one or more files;

wherein the one or more files are organized as a version set by the file server indicating a set of physical blocks that are shared among the one or more files;

wherein the one or more files organized as the version set represents the logical disk storage for respective virtual machines of a group of one or more virtual machines in the storage virtualization environment; and

wherein the version set is a sparse snapshot of the file of the one or more files.

18. A computer program product for performing an I/O on a storage object of one or more storage objects, the computer program product operating in a storage virtualization environment that includes a storage server, a data storage system and a client of the storage server, wherein the data storage system in communication with the storage server provides a pool of storage resources to the storage server for storing the one or more storage objects as a set of virtual devices that share from the pool of storage resources organized as a set of related objects indicating a portion of data that is shared among the one or more storage objects, wherein the storage object of the one or more storage objects is presented to an application running in the client as a logically addressable storage object, wherein the computer program product includes computer-executable logic encoded on a non-transitory computer-readable medium for executing the following steps:

retrieving from a mapping client a physical address for a logical addressable storage object on which I/O is generated by an application, wherein the mapping client resides in the client of the storage server and provides the application a mapping of the logical addressable storage object to its physical location on the data storage system;

examining a data cache included in the client of the storage server to determine the availability of the storage object in the data cache, wherein the data cache contains the physical address and content of the one or more storage objects accessed by previous I/O operations; and

completing the I/O on finding the storage object in the data cache;

wherein the one or more storage objects is represented by one or more files;

wherein the storage server is represented by a file server in communication with the data storage system;

wherein the data storage system provides the logical disk storage to the file server for storing the one or more files;

wherein the one or more files are organized as a version set by the file server indicating a set of physical blocks that are shared among the one or more files;

wherein the one or more files organized as the version set represents the logical disk storage for respective virtual machines of a group of one or more virtual machines in the storage virtualization environment; and

wherein the version set is a sparse snapshot of the file of the one or more files.