System and method for assigning memory reserved for high availability failover to virtual machines

• US 9,830,236 B2
• Filed: 06/05/2013
• Issued: 11/28/2017
• Est. Priority Date: 06/05/2013
• Status: Active Grant

• Current Assignee: Vmware LLC (Broadcom, Inc.)
• Original Assignee: VMware, Inc. (Broadcom, Inc.)
• Inventors: Antony, Jinto
• Primary Examiner(s): Kim, Matt
• Assistant Examiner(s): CONTINO, PAUL F
• Application Number:
  US13/910,141
• Publication Number:
  US 20140365816A1
• Time in Patent Office: 1,637 Days
• CPC Class Codes: G06F 11/1484 involving virtual machines
• CPC Class Codes: G06F 11/203 using migration
• CPC Class Codes: G06F 11/2033 switching over of hardware resources
• CPC Class Codes: G06F 11/2043 where the redundant components share a common memory address space
• CPC Class Codes: G06F 11/2046 where the redundant components share persistent storage G06F11/2043 takes precedence
• CPC Class Codes: G06F 2009/45583 Memory management, e.g. access or allocation
• CPC Class Codes: G06F 9/45558 Hypervisor-specific management and integration aspects

Abstract

Techniques for assigning memory reserved for high availability (HA) failover to virtual machines in high availability (HA) enabled clusters are described. In one embodiment, the memory reserved for HA failover is determined in each host computing system of the HA cluster. Further, the memory reserved for HA failover is assigned to one or more virtual machines in the HA cluster as input/output (I/O) cache memory at a first level.

Claims

1. A method for assigning memory reserved for high availability (HA) failover, comprising:

determining the memory reserved for HA failover in a first host computing system of a HA cluster, wherein the memory reserved for HA failover is memory that is reserved in the first host computing system in the HA cluster to accommodate affected virtual machines in an event of HA failover, wherein the HA failover is a failure of any host computing system in the HA cluster;

determining a first virtual machine, which requires input/output (I/O) cache memory to improve I/O performance, running in a second host computing system of the HA cluster;

when there is no failure of any host computing system in the HA cluster, remotely assigning, at a first level, the memory reserved for HA failover in the first host computing system to the first virtual machine running on the second host computing system as I/O cache memory for carrying, out read and write operations to improve the I/O performance of the first virtual machine; and

when there is a failure of a host computing system in the HA cluster, reclaiming, by the first host computing system, the memory reserved for HA failover that is remotely assigned as the I/O cache memory from the first virtual machine to accommodate virtual machines of the failed host computing system.

2. The method of claim 1, further comprising:

repeating the steps of claim 1 for each virtual machine requiring the I/O cache memory.

3. The method of claim 1, wherein remotely assigning the memory reserved for HA failover to the first virtual machine as the I/O cache memory, comprises one of:

assigning the memory reserved for HA failover remotely to the first virtual machine as the I/O cache memory using a connection between the first host computing system having the memory reserved for HA failover and the second host computing system; and

live migrating the first virtual machine to the first host computing system having the memory reserved for HA failover and assigning the memory reserved for HA failover locally to the migrated first virtual machine as the I/O cache memory.

4. The method of claim 3, wherein remotely assigning the memory, reserved for HA failover to the first virtual machine as the I/O cache memory using the connection between the first host computing system having the memory reserved for HA failover and the second host computing system, comprises:

assigning the memory reserved for HA failover remotely to the first virtual machine as the I/O cache memory using a remote direct memory access (RDMA)-enabled network connection between the first host computing system and the second host computing system.

5. The method of claim 1, further comprising:

assigning a first portion of the memory reserved for HA failover locally to a second virtual machine running on the first host computing system as at least one of the I/O cache memory and a virtual machine memory; and

assigning a second portion of the memory reserved for HA failover remotely to the first virtual machine running on the second host computing system as the I/O cache memory.

6. The method of claim 5, wherein the memory reserved for HA failover is assigned to the second virtual machine in the HA cluster between the I/O cache memory and the virtual machine memory using a resource scheduling framework.

7. The method of claim 5, further comprising:

reclaiming the I/O cache memory and the virtual machine memory assigned to the second virtual machine by the first host computing system when there is a failure of any host computing system in the HA cluster.

8. The method of claim 7, wherein reclaiming the virtual machine memory comprises:

reclaiming the virtual machine memory from the second virtual machine by the first host computing system using a memory ballooning mechanism.

9. The method of claim 7 , wherein reclaiming the I/O cache memory comprises:

discarding the I/O cache memory associated with the first and second virtual machines after committing pending write operations from the I/O cache memory to a persistent storage; and

reclaiming the memory reserved for HA failover from the first and second virtual machines by the first host computing system upon discarding the I/O cache memory.

10. The method of claim 1, further comprising:

assigning one or more solid-state drives (SSDs) or a portion of the one or more SSDs to the first virtual machine in the HA cluster as the I/O cache memory at a second level.

11. A system comprising:

one or more high availability (HA) clusters, each HA cluster including a plurality of host computing systems with each host computing system including virtual machines running therein; and

a management server communicative with the one or more HA clusters, wherein the management server includes a memory management module to;

determine memory reserved for HA failover in a first host computing system of a HA cluster of the one or more HA clusters, wherein the memory reserved for HA failover is memory that is reserved in the first host computing system in the HA cluster to accommodate affected virtual machines in an event of HA failover, wherein the HA failover is a failure of any host computing system in the HA cluster;

determine a first virtual machine, which requires input/output (I/O) cache memory to improve I/O performance, running in a second host computing system of the HA cluster;

when there is no failure of any host computing system in the HA cluster, remotely assign, at a first level, the memory reserved for HA failover in the first host computing system to the first virtual machine running on the second host computing system as I/O cache memory for carrying out read and write operations to improve the I/O performance of the first virtual machine; and

when there is a failure of a host computing system in the HA cluster, enable the first host computing system to reclaim the memory reserved for HA failover that is remotely assigned as the I/O cache memory from the first virtual machine to accommodate affected virtual machines of the failed host computing system.

12. The system of claim 11, wherein the memory management module is further configured to:

repeat the steps of claim 11 for each virtual machine requiring the I/O cache memory.

13. The system of claim 11, wherein the memory management module remotely assigns the memory reserved for HA failover to the first virtual machine as the I/O cache memory using a connection between the first host computing system having the memory reserved for HA failover and the second host computing system.

14. The system of claim 13, wherein the memory management module remotely assigns the memory reserved for HA failover to the first virtual machine as the I/O cache memory using a remote direct memory access (RDMA)-enabled network connection between the first host computing system and the second host computing system.

15. The system of claim 11, wherein the memory management module live migrates the first virtual machine to the first host computing system having the memory reserved for HA failover and assigns the memory reserved for HA fallover locally to the migrated first virtual machine as the I/O cache memory.

16. The system of claim 11, wherein the memory management module assigns a first portion of the memory reserved for HA failover locally to a second virtual machine running on the first host computing system as at least one of the I/O cache memory and a virtual machine memory, and assigns a second portion of the memory reserved for HA failover remotely to the first virtual machine running on the second host computing system as the I/O cache memory.

17. The system of claim 16, wherein the memory management module enables the first host computing system to reclaim the I/O cache memory and the virtual machine memory assigned to the second virtual machine in the event of the HA failover in the HA cluster.

18. The system of claim 11, further comprising:

one or more solid-state drives (SSDs), wherein the memory management module assigns the one or more SSDs or a portion of the one or more SSDs to the first virtual machine in the HA cluster as the I/O cache memory at a second level.

19. A non-transitory computer-readable storage medium including instructions that are configured, when executed by a computing system, to perform a method for assigning memory reserved for high availability (HA) failover, the method comprising:

determining the memory reserved for HA failover in a first host computing system of a HA cluster, wherein the memory reserved for HA failover is memory that is reserved in the first host computing system in the HA cluster to accommodate affected virtual machines in an event of HA failover, wherein the HA failover is a failure of any host computing system in the HA cluster;

determining a first virtual machine, which requires input/output (I/O) cache memory to improve (I/O) performance, running in a second host computing system of the HA cluster;

when there is no failure of any host computing system of the HA cluster, remotely assigning, at a first level, the memory reserved for HA failover in the first host computing system to the first virtual machine running on the second host computing system as I/O cache memory for carrying out read and write operations to improve the I/O performance of the first virtual machine; and

when there is a failure of a host computing system in the HA cluster, enabling the first host computing system to reclaim the memory reserved for HA failover that is remotely assigned as the I/O cache memory from the first virtual machine to accommodate affected virtual machines of the failed host computing system.

20. The non-transitory computer-readable storage medium of claim 19, further comprising:

repeating the steps of claim 19 for each virtual machine requiring the I/O cache memory.

21. The non-transitory computer-readable storage medium of claim 19, wherein remotely assigning the memory reserved for HA failover to the first virtual machine as the I/O cache memory, comprises one of:

assigning the memory reserved for HA failover remotely to the first virtual machine as the I/O cache memory using a connection between the first host computing system having the memory reserved for HA failover and the second host computing system; and

live migrating the first virtual machine to the first host computing system having the memory reserved for HA failover and assigning the memory reserved for HA failover locally to the migrated first virtual machine as the I/O cache memory.

22. The non-transitory computer-readable storage medium of claim 21, wherein remotely assigning the memory reserved for HA failover to the first virtual machine as the I/O cache memory using the connection between the first host computing system having the memory reserved for HA failover and the second host computing system, comprises:

assigning the memory reserved for HA failover remotely to the first virtual machine as the I/O cache memory using a remote direct memory access (RDMA)-enabled network connection between the first host computing system and the second host computing system.

23. The non-transitory computer-readable storage medium of claim 19, further comprising instructions to:

assign a first portion of the memory reserved for HA failover locally to a second virtual machine running on the first host computing system as at least one of the I/O cache memory and a virtual machine memory; and

assigning a second portion of the memory reserved for HA failover remotely to the first virtual machine running on the second host computing system as the I/O cache memory.

24. The non-transitory computer-readable storage medium of claim 23, wherein the memory reserved for HA failover is assigned to the second virtual machine in the HA cluster between the I/O cache memory and the virtual machine memory using a resource scheduling framework.

25. The non-transitory computer-readable storage medium of claim 23, further comprising:

reclaiming the I/O cache memory and the virtual machine memory assigned to the second virtual machine by the first host computing system in the event of the HA failover in the HA cluster.

26. The non-transitory computer-readable storage medium of claim 25, wherein reclaiming the I/O cache memory comprises:

discarding the I/O cache memory associated with the first and second virtual machines after committing pending write operations from the I/O cache memory to a persistent storage; and

reclaiming the memory reserved for HA failover from the first and second virtual machines by the first host computing system upon discarding the I/O cache memory.

27. The non-transitory computer-readable storage medium of claim 19, further comprising:

assigning one or more solid-state drives (SSDs) or a portion of the one or more SSDs to the first virtual machine in the HA cluster as the I/O cache memory at a second level.