Method and system for scheduling virtual machines in integrated virtual machine clusters

• US 10,031,777 B2
• Filed: 11/13/2015
• Issued: 07/24/2018
• Est. Priority Date: 11/14/2014
• Status: Active Grant

• Current Assignee: Sangfor Technologies, Inc.
• Original Assignee: Sangfor Technologies, Inc.
• Inventors: Wang, Zheng
• Primary Examiner(s): Patel, Hiren
• Application Number:
  US14/940,286
• Publication Number:
  US 20160139957A1
• Time in Patent Office: 984 Days
• Field of Search: None
• CPC Class Codes: G06F 2009/4557 Distribution of virtual machine instances; Migration and load balancing
• CPC Class Codes: G06F 9/45554 Instruction set architectures of guest OS and hypervisor or native processor differ, e.g. Bochs or VirtualPC on PowerPC MacOS
• CPC Class Codes: G06F 9/45558 Hypervisor-specific management and integration aspects
• CPC Class Codes: G06F 9/50 Allocation of resources, e.g. of the central processing unit [CPU]
• CPC Class Codes: G06F 9/5077 Logical partitioning of resources; Management or configuration of virtualized resources specific details on emulation or internal functioning of virtual machines G06F9/455

Abstract

A method for scheduling virtual machines in a virtual machine cluster includes obtaining a filename of a target virtual machine when a user requests to start the target virtual machine; inquiring, based on the filename of the target virtual machine, a storage module or a database to acquire one or more nodes where copies of the target virtual machine are located; selecting, from the acquired one or more nodes, a node with a highest score as a target node having a copy of the target virtual machine; and running the copy of the target virtual machine on the selected target node with the highest score.

Claims

1. A method for scheduling virtual machines in an integrated virtual machine cluster, the method being performed by a computer program executable by a processor, the method comprising:

S1). obtaining a filename of a target virtual machine from a directory created for the virtual machines when a user requests to start the target virtual machine;

S2). inquiring, based on the filename of the target virtual machine, a storage module or a database to acquire nodes where copies of the target virtual machine are located, wherein a node is a physical machine or computer for hosting one or more virtual machines;

S3). selecting, based on a relationship between the acquired nodes where copies of the target virtual machine are located and nodes of the integrated virtual machine cluster that meet a CPU and memory configuration requirement of the target virtual machine, a set of nodes for starting the target virtual machine, and calculating a score for each node in the selected set of nodes to determine a target node; and

S4). running the target virtual machine on the determined target node,

wherein (S3) further includes;

S30). obtaining configurations of CPU and memory required by the target virtual machine,

S31). locating all nodes in the integrated virtual machine cluster to obtain a first set of nodes,

S32). judging whether each node in the first set of nodes meets the CPU and memory configuration requirement of the target virtual machine, and removing nodes in the first set of nodes that do not meet the configuration requirement, so as to create a second set of nodes,

S33). determining the acquired nodes in (S2) as a third set of nodes,

S34). in response to an intersection existing between the second set of nodes and the third set of nodes, determining the intersection of the second set of nodes and the third set of nodes as a fourth set of nodes,

S35). calculating a score of each node in the fourth set of nodes, and

S36). selecting a node with a highest score as the target node, wherein the highest score indicating a highest efficiency for running the target virtual machine on the selected target node.

2. The method according to claim 1 between (S33) and (S34), further including:

judging whether there is intersection between the second set of nodes and the third set of nodes;

when there is no intersection between the second set of nodes and the third set of nodes, determining the second set of nodes as the fourth set of nodes, continuing to (S35); and

when there is intersection between the second set of nodes and the third set of nodes, continuing to (S34).

3. The method according to claim 2, after (S4), further including:

judging whether the selected target node belongs to the third set of nodes; and

when the selected target node does not belong to the third set of nodes, continuing to (S3).

4. The method according to claim 1, between (S32) and (S33), further including:

judging whether the second set of nodes is empty; and

when the second set of nodes is empty, prompting a boot failure.

5. The method according to claim 1, wherein:

provided that S is the score of a current node;

Fo is CPU frequency of the current node;

F is a sum of the CPU frequency of all nodes in the integrated virtual machine cluster;

N is a number of virtual machines in the integrated virtual machine cluster; and

No is a number of virtual machines running on the current node;

the score of each node in the fourth set of nodes is calculated by;

S=Fo/F*(N+1)−

No.

6. The method according to claim 1, wherein (S35) further includes:

calculating the score of each node in the fourth set of nodes based on parameters including a number of virtual machines running on the node.

7. A system for scheduling virtual machines in an integrated virtual machine cluster, comprising at least one processor, the at least one processor being configured to:

acquire a filename of a target virtual machine from a directory created for the virtual machines when a user requests to start the target virtual machine;

based on the filename of the target virtual machine, inquire a storage module or a database to acquire nodes where copies of the target virtual machine are located, wherein a node is a physical machine or computer for hosting one or more virtual machines;

based on a relationship between the acquired nodes where copies of the target virtual machine are located and nodes of the integrated virtual machine cluster that meet a CPU and memory configuration requirement of the target virtual machine, select a set of nodes for starting the target virtual machine;

calculating a score for each node in the selected set of nodes to determine a target node; and

run the target virtual machine on the determined target node;

wherein further includes;

obtaining configurations of CPU and memory required by the target virtual machine;

locating all nodes in the integrated virtual machine cluster to obtain a first set of nodes;

judging whether each node in the first set of nodes meets the CPU and memory configuration requirement of the target virtual machine, and removing nodes in the first set of nodes that do not meet the configuration requirement, so as to create a second set of nodes;

determining the acquired nodes in (S2) as a third set of nodes;

in response to an intersection existing between the second set of nodes and the third set of nodes, determining the intersection of the second set of nodes and the third set of nodes as a fourth set of nodes;

calculating a score of each node in the fourth set of nodes; and

selecting a node with a highest score as the target node, wherein the highest score indicating a highest efficiency for running the target virtual machine on the selected target node.

8. The system according to claim 7, wherein the at least one processor is further configured to:

determine whether there is intersection between the second set of nodes and the third set of nodes; and

when there is intersection between the second set of nodes and the third set of nodes, determine the intersection between the second set of nodes and the third set of nodes as a fourth set of nodes and, when there is no intersection between the second set of nodes and the third set of nodes, determine the second set of nodes as the fourth set of nodes.

9. The system according to claim 8, wherein the at least one processor is further configured to, after running the target virtual machine on the selected target node, judge whether the selected target node belongs to the third set of nodes, and when the selected target node does not belong to the third set of nodes, re-select a set of nodes for starting the target virtual machine based on the relationship between the acquired nodes where copies of the target virtual machine are located and the nodes of the integrated virtual machines that meet the CPU and memory configuration requirement of the target virtual machine, and calculate a score for each node in the re-selected set of nodes to determine another target node as the target node.

10. The system according to claim 7, wherein the at least one processor is further configured to:

determine whether the second set of nodes is empty; and

when the second set of nodes is empty, prompt a boot failure.

11. The system according to claim 7, wherein:

provided that S is the score of a current node;

Fo is CPU frequency of the current node;

F is a sum of the CPU frequency of all nodes in the integrated virtual machine cluster;

N is a number of virtual machines in the integrated virtual machine cluster; and

No is a number of virtual machines running on the current node;

the at least one processor is further configured to calculate the score of each node in the fourth set of nodes by;

S=Fo/F*(N+1)−

No.

12. A non-transitory computer-readable medium having computer program for, when being executed by a processor, performing a method for scheduling virtual machines in an integrated virtual machine cluster, the method comprising:

S1). obtaining a filename of a target virtual machine from a directory created for the virtual machines when a user requests to start the target virtual machine;

S2). inquiring, based on the filename of the target virtual machine, a storage module or a database to acquire nodes where copies of the target virtual machine are located, wherein a node is a physical machine or computer for hosting one or more virtual machines;

S3). selecting, based on a relationship between the acquired nodes where copies of the target virtual machine are located and nodes of the integrated virtual machine cluster that meet a CPU and memory configuration requirement of the target virtual machine, a set of nodes for starting the target virtual machine, and calculating a score for each node in the selected set of nodes to determine a target node; and

S4). running the target virtual machine on the determined target node,

wherein (S3) further includes;

S30). obtaining configurations of CPU and memory required by the target virtual machine,

S31). locating all nodes in the integrated virtual machine cluster to obtain a first set of nodes,

S32). judging whether each node in the first set of nodes meets the CPU and memory configuration requirement of the target virtual machine, and removing nodes in the first set of nodes that do not meet the configuration requirement, so as to create a second set of nodes,

S33). determining the acquired nodes in (S2) as a third set of nodes,

S34). in response to an intersection existing between the second set of nodes and the third set of nodes, determining the intersection of the second set of nodes and the third set of nodes as a fourth set of nodes,

S35). calculating a score of each node in the fourth set of nodes, and

36). selecting a node with a highest score as the target node, wherein the highest score indicating a highest efficiency for running the target virtual machine on the selected target node.

13. The non-transitory computer-readable medium according to claim 12, between (S33) and (S34), the method further including:

judging whether there is intersection between the second set of nodes and the third set of nodes;

when there is no intersection between the second set of nodes and the third set of nodes, determining the second set of nodes as the fourth set of nodes, continuing to (S35); and

when there is intersection between the second set of nodes and the third set of nodes, continuing to (S34).

14. The non-transitory computer-readable medium according to claim 13, after (S4), the method further including:

judging whether the selected target node belongs to the third set of nodes; and

when the selected target node does not belong to the third set of nodes, continuing to (S3).

15. The non-transitory computer-readable medium according to claim 12, between (S32) and (S33), the method further including:

judging whether the second set of nodes is empty; and

when the second set of nodes is empty, prompting a boot failure.

16. The non-transitory computer-readable medium according to claim 12, wherein:

provided that S is the score of a current node;

Fo is CPU frequency of the current node;

F is a sum of the CPU frequency of all nodes in the integrated virtual machine cluster;

N is a number of virtual machines in the integrated virtual machine cluster; and

No is a number of virtual machines running on the current node;

the score of each node in the fourth set of nodes is calculated by;

S=Fo/F*(N+1)−

No.