Method, apparatus and computer programs providing cluster-wide page management

• US 10,007,609 B2
• Filed: 03/01/2016
• Issued: 06/26/2018
• Est. Priority Date: 01/16/2013
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Dube, Parijat, Guerin, Xavier R., Seelam, Seetharami R.
• Primary Examiner(s): Avellino, Joseph E
• Assistant Examiner(s): MCLEOD, MARSHALL M
• Application Number:
  US15/057,259
• Publication Number:
  US 20160179669A1
• Time in Patent Office: 847 Days
• Field of Search: 709212, 709213, 709217, 711209, 711141
• CPC Class Codes: G06F 12/08 in hierarchically structured memory systems, e.g. virtual memory systems
• CPC Class Codes: G06F 12/084 with a shared cache
• CPC Class Codes: G06F 12/0864 using pseudo-associative means, e.g. set-associative or hashing
• CPC Class Codes: G06F 16/00 Information retrieval; Database structures therefor; File system structures therefor
• CPC Class Codes: G06F 2009/45583 Memory management, e.g. access or allocation
• CPC Class Codes: G06F 2212/152 Virtualized environment, e.g. logically partitioned system
• CPC Class Codes: G06F 2212/163 Server or database system
• CPC Class Codes: G06F 2212/62 Details of cache specific to multiprocessor cache arrangements
• CPC Class Codes: G06F 9/45558 Hypervisor-specific management and integration aspects

Abstract

A data processing system includes a plurality of virtual machines each having associated memory pages; a shared memory page cache that is accessible by each of the plurality of virtual machines; and a global hash map that is accessible by each of the plurality of virtual machines. The data processing system is configured such that, for a particular memory page stored in the shared memory page cache that is associated with two or more of the plurality of virtual machines, there is a single key stored in the global hash map that identifies at least a storage location in the shared memory page cache of the particular memory page. The system can be embodied at least partially in a cloud computing system.

Claims

1. A data processing system comprising:

at least one data processor; and

at least one non-transitory memory including computer program code, wherein the at least one processor, in response to execution of the computer program code, is configured to comprise;

a plurality of virtual machines each having associated memory pages;

a shared memory page cache that is accessible by each of the plurality of virtual machines; and

a global hash map that is accessible by each of the plurality of virtual machines;

where for a particular memory page stored in the shared memory page cache that is accessed by two or more of the plurality of virtual machines there is a single key stored in the global hash map that uniquely identifies at least a storage location in the shared memory page cache of the particular memory page enabling page de-duplication to be achieved.

2. The system as in claim 1, where the key is comprised at least in part of a result of processing the particular memory page using a hashing algorithm.

3. The system of claim 1, where the shared memory page cache is comprised of an in-memory shared memory page cache and where said virtual machines each comprise a remote direct memory access (RDMA) adapter for accessing the in-memory shared memory page cache.

4. The system of claim 1, where a virtual machine is configured, prior to migrating the virtual machine from a first instance of a virtual machine to a second instance of a virtual machine, to swap out associated memory pages to the shared memory page cache and to send information relating to a corresponding plurality of keys to the second instance of the virtual machine enabling the second instance of the virtual machine to swap-in a corresponding plurality of the memory pages from the shared memory page cache.

5. The system as in claim 4, where the information comprises a pointer to a location where the plurality of keys are stored.

6. The system as in claim 5, where the plurality of keys are stored with the memory pages in the shared memory page cache.

7. The system as in claim 1, where a hypervisor of a virtual machine is configured, prior to migrating the virtual machine from a first instance of a virtual machine to a second instance of a virtual machine, to swap-out all associated memory pages to the shared memory page cache while computing a key for each of the associated memory pages.

8. The system as in claim 7, where the hypervisor is further configured to determine if a computed key is already present in the global hash map and, only if it is determined that the computed key is not already present in the global hash map, to store the computed key in the global hash map and the instance of the memory page in the shared memory page cache.

9. The system as in claim 8, where the hypervisor is further configured to send information relating to the computed keys to the second instance of the virtual machine to enable the second instance of the virtual machine to swap-in a corresponding plurality of the memory pages from the shared memory page cache.

10. The system as in claim 1, where a system server that is responsible for one or both of the shared memory page cache and the global hash map is replicated so as to provide fault tolerance against a failure of the system server.

11. The system as in claim 1, embodied at least partially in a cloud computing system.

12. A data processing system comprising:

at least one data processor; and

at least one non-transitory memory including computer program code, wherein the at least one processor, in response to execution of the computer program code, is configured to comprise;

a plurality of virtual machines each having associated memory pages;

a shared memory page cache that is accessible by each of the plurality of virtual machines; and

a global hash map that is accessible by each of the plurality of virtual machines;

where for a particular memory page stored in the shared memory page cache that is accessed by two or more of the plurality of virtual machines there is a single key stored in the global hash map that uniquely identifies at least a storage location in the shared memory page cache of the particular memory page enabling page de-duplication to be achieved; and

where a hypervisor of a virtual machine is configured, prior to migrating the virtual machine from a first instance of a virtual machine to a second instance of a virtual machine, to swap-out all associated memory pages to the shared memory page cache while computing a key for each of the associated memory pages.

13. The system as in claim 12, where the hypervisor is further configured to determine if a computed key is already present in the global hash map and, only if it is determined that the computed key is not already present in the global hash map, to store the computed key in the global hash map and the instance of the memory page in the shared memory page cache.

14. The system as in claim 13, where the hypervisor is further configured to send information relating to the computed keys to the second instance of the virtual machine to enable the second instance of the virtual machine to swap-in a corresponding plurality of the memory pages from the shared memory page cache.

15. The system as in claim 12, where a virtual machine is configured, prior to migrating the virtual machine from a first instance of a virtual machine to a second instance of a virtual machine, to swap out associated memory pages to the shared memory page cache and to send information relating to a corresponding plurality of keys to the second instance of the virtual machine enabling the second instance of the virtual machine to swap-in a corresponding plurality of the memory pages from the shared memory page cache.

16. The system as in claim 12, where the key is comprised at least in part of a result of processing the particular memory page using a hashing algorithm.

17. The system of claim 12, where the shared memory page cache is comprised of an in-memory shared memory page cache and where said virtual machines each comprise a remote direct memory access (RDMA) adapter for accessing the in-memory shared memory page cache.

18. The system as in claim 15, where the information comprises a pointer to a location where the plurality of keys are stored.

19. The system as in claim 18, where the plurality of keys are stored with the memory pages in the shared memory page cache.

20. The system as in claim 12, where a system server that is responsible for one or both of the shared memory page cache and the global hash map is replicated so as to provide fault tolerance against a failure of the system server.