NUMA-based client placement

• US 10,248,977 B2
• Filed: 04/09/2014
• Issued: 04/02/2019
• Est. Priority Date: 08/24/2013
• Status: Active Grant

• Current Assignee: Vmware LLC (Broadcom, Inc.)
• Original Assignee: VMware, Inc. (Broadcom, Inc.)
• Inventors: Parikh, Aashish, Zaroo, Puneet, Shanmuganathan, Ganesha
• Primary Examiner(s): Anyan, Barbara B
• Application Number:
  US14/249,267
• Publication Number:
  US 20150058400A1
• Time in Patent Office: 1,819 Days
• CPC Class Codes: G06F 1/26 Power supply means, e.g. regulation thereof for memories G11C
• CPC Class Codes: G06F 1/3209 Monitoring remote activity, e.g. over telephone lines or network connections
• CPC Class Codes: G06F 11/3062 where the monitored property is the power consumption power management in a computing system G06F1/3203
• CPC Class Codes: G06F 9/5044 considering hardware capabilities
• CPC Class Codes: G06F 9/505 considering the load
• 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
• CPC Class Codes: G06F 9/5094 where the allocation takes into account power or heat criteria power management in computers in general G06F1/3203; thermal management in computers in general G06F1/206
• CPC Class Codes: G06Q 30/0283 Price estimation or determination
• CPC Class Codes: G06Q 50/06 Energy or water supply
• CPC Class Codes: H04L 67/01 Protocols
• CPC Class Codes: H04L 67/1001 for accessing one among a plurality of replicated servers
• 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]
• CPC Class Codes: Y02D 10/00 Energy efficient computing, e.g. low power processors, power management or thermal management

Abstract

A management server and method for performing resource management operations in a distributed computer system takes into account information regarding multi-processor memory architectures of host computers of the distributed computer system, including information regarding Non-Uniform Memory Access (NUMA) architectures of at least some of the host computers, to make a placement recommendation to place a client in one of the host computers.

Claims

1. A method for performing resource management operations in a distributed computer system, the method comprising:

receiving, from each of host computers of the distributed computer system, information regarding multi-processor memory architectures of the host computer, including information regarding Non-Uniform Memory Access (NUMA) architectures of the host computer;

receiving, from said each of the host computers of the distributed computer system, information regarding physical and virtual processors operating in the host computer;

receiving, from a user, virtual processor requirements of a client to be placed in one of the host computers of the distributed computer system; and

recommending a target host computer of the distributed computer system to place the client, taking into account the received information regarding the multi-processor memory architectures of the host computers of the distributed computer system, the received information regarding the physical and virtual processors in the host computers of the distributed computer system and the received virtual processor requirements of the client to be placed.

2. The method of claim 1, wherein the recommending the target host computer includes computing NUMA scores for the host computers of the distributed computer system, wherein at least one NUMA score is derived using virtual and physical processor ratios, each ratio being the number of virtual processors associated with a NUMA node of one of the host computers over the number of physical processors associated with that NUMA node.

3. The method of claim 2, wherein at least one of the virtual and physical processor ratios for a particular host computer includes the number of virtual processors of the client to be placed.

4. The method of claim 2, wherein the computing the NUMA scores for the host computers of the distributed computer system includes calculating a statistical measure of the virtual and physical processor ratios for a particular host computer to compute the at least one NUMA score for the particular host computer.

5. The method of claim 4, wherein the computing the NUMA scores for the host computers of the distributed computer system includes selecting a maximum value from the virtual and physical processor ratios for a particular host computer.

6. The method of claim 4, wherein the computing the NUMA scores for the host computers of the distributed computer system includes computing a vector magnitude of a vector of the virtual and physical processor ratios for a particular host computer.

7. The method of claim 4, wherein the computing the NUMA scores for the host computers of the distributed computer system includes computing a k-th percentile of the virtual and physical processor ratios for a particular host computer, where k is a number less than 100.

8. The method of claim 4, wherein the computing the NUMA scores for the host computers of the distributed computer system includes computing a standard deviation or Jain'"'"'s fairness index of the virtual and physical processor ratios for a particular host computer.

9. The method of claim 1, wherein the recommending of the target host computer is part of an initial placement operation to initially place the client in the distributed computer system.

10. The method of claim 1, wherein the recommending of the target host computer is part of a load balancing operation to migrate the client from an original host computer in the distributed computer system to the target host computer.

11. A non-transitory computer-readable storage medium containing program instructions for performing resource management operations in a distributed computer system, wherein execution of the program instructions by one or more processors of a computer system causes the one or more processors to perform a method comprising:

receiving, from each of host computers of the distributed computer system, information regarding multi-processor memory architectures of the host computer, including information regarding Non-Uniform Memory Access (NUMA) architectures of the host computer;

receiving, from said each of the host computers of the distributed computer system, information regarding physical and virtual processors operating in the host computer;

receiving, from a user, virtual processor requirements of a client to be placed in one of the host computers of the distributed computer system; and

recommending a target host computer of the distributed computer system to place the client, taking into account the received information regarding the multi-processor memory architectures of the host computers of the distributed computer system, the received information regarding the physical and virtual processors in the host computers of the distributed computer system and the received virtual processor requirements of the client to be placed.

12. The non-transitory computer-readable storage medium of claim 11, wherein the recommending the target host computer includes computing NUMA scores for the host computers of the distributed computer system, wherein at least one NUMA score is derived using virtual and physical processor ratios, each ratio being the number of virtual processors associated with a NUMA node of one of the host computers over the number of physical processors associated with that NUMA node.

13. The non-transitory computer-readable storage medium of claim 12, wherein at least one of the virtual and physical processor ratios for a particular host computer includes the number of virtual processors of the client to be placed.

14. The non-transitory computer-readable storage medium of claim 12, wherein the computing the NUMA scores for the host computers of the distributed computer system includes calculating a statistical measure of the virtual and physical processor ratios for a particular host computer to compute the at least one NUMA score for the particular host computer.

15. The non-transitory computer-readable storage medium of claim 14, wherein the computing the NUMA scores for the host computers of the distributed computer system includes selecting a maximum value from the virtual and physical processor ratios for a particular host computer.

16. The non-transitory computer-readable storage medium of claim 14, wherein the computing the NUMA scores for the host computers of the distributed computer system includes computing a vector magnitude of a vector of the virtual and physical processor ratios for a particular host computer.

17. The non-transitory computer-readable storage medium of claim 14, wherein the computing the NUMA scores for the host computers of the distributed computer system includes computing a k-th percentile of the virtual and physical processor ratios for a particular host computer, where k is a number less than 100.

18. The non-transitory computer-readable storage medium of claim 14, wherein the computing the NUMA scores for the host computers of the distributed computer system includes computing a standard deviation or Jain'"'"'s fairness index of the virtual and physical processor ratios for a particular host computer.

19. The non-transitory computer-readable storage medium of claim 11, wherein the recommending of the target host computer is part of an initial placement operation to initially place the client in the distributed computer system.

20. The non-transitory computer-readable storage medium of claim 11, wherein the recommending of the target host computer is part of a load balancing operation to migrate the client from an original host computer in the distributed computer system to the target host computer.

21. A management server for a distributed computer system comprising:

memory; and

a processor configured to;

receive, from each of host computers of the distributed computer system, information regarding multi-processor memory architectures of the host computer, including information regarding Non-Uniform Memory Access (NUMA) architectures of the host computer;

receive, from said each of the host computers of the distributed computer system, information regarding physical and virtual processors operating in the host computer;

receive, from a user, virtual processor requirements of a client to be placed in one of the host computers of the distributed computer system; and

recommend a target host computer of the distributed computer system to place the client, taking into account the received information regarding the multi-processor memory architectures of the host computers of the distributed computer system, the received information regarding the physical and virtual processors in the host computers of the distributed computer system, and the received virtual processor requirements of the client to be placed.

22. The management server of claim 21, wherein the processor is configured to compute NUMA scores for the host computers of the distributed computer system, wherein at least one NUMA score is derived using virtual and physical processor ratios, each ratio being the number of virtual processors associated with a NUMA node of one of the host computers over the number of physical processors associated with that NUMA node.

23. The management server of claim 22, wherein at least one of the virtual and physical processor ratios for a particular host computer used by the processor includes the number of virtual processors of the client to be placed.

24. The management server of claim 22, wherein the processor is configured to calculate a statistical measure of the virtual and physical processor ratios for a particular host computer to compute the at least one NUMA score for the particular host computer.