WO2010111151A2 - Virtual non-uniform memory architecture for virtual machines - Google Patents
Virtual non-uniform memory architecture for virtual machines Download PDFInfo
- Publication number
- WO2010111151A2 WO2010111151A2 PCT/US2010/028038 US2010028038W WO2010111151A2 WO 2010111151 A2 WO2010111151 A2 WO 2010111151A2 US 2010028038 W US2010028038 W US 2010028038W WO 2010111151 A2 WO2010111151 A2 WO 2010111151A2
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- virtual
- numa
- computer system
- virtual machine
- numa node
- Prior art date
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5061—Partitioning or combining of resources
- G06F9/5077—Logical partitioning of resources; Management or configuration of virtualized resources
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0602—Interfaces specially adapted for storage systems specifically adapted to achieve a particular effect
- G06F3/0604—Improving or facilitating administration, e.g. storage management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0628—Interfaces specially adapted for storage systems making use of a particular technique
- G06F3/0629—Configuration or reconfiguration of storage systems
- G06F3/0631—Configuration or reconfiguration of storage systems by allocating resources to storage systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0628—Interfaces specially adapted for storage systems making use of a particular technique
- G06F3/0662—Virtualisation aspects
- G06F3/0665—Virtualisation aspects at area level, e.g. provisioning of virtual or logical volumes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0668—Interfaces specially adapted for storage systems adopting a particular infrastructure
- G06F3/0671—In-line storage system
- G06F3/0683—Plurality of storage devices
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45504—Abstract machines for programme code execution, e.g. Java virtual machine [JVM], interpreters, emulators
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/48—Program initiating; Program switching, e.g. by interrupt
- G06F9/4806—Task transfer initiation or dispatching
- G06F9/4843—Task transfer initiation or dispatching by program, e.g. task dispatcher, supervisor, operating system
- G06F9/485—Task life-cycle, e.g. stopping, restarting, resuming execution
- G06F9/4856—Task life-cycle, e.g. stopping, restarting, resuming execution resumption being on a different machine, e.g. task migration, virtual machine migration
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F12/00—Accessing, addressing or allocating within memory systems or architectures
- G06F12/02—Addressing or allocation; Relocation
- G06F12/0223—User address space allocation, e.g. contiguous or non contiguous base addressing
- G06F12/0284—Multiple user address space allocation, e.g. using different base addresses
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
- G06F2009/45562—Creating, deleting, cloning virtual machine instances
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2212/00—Indexing scheme relating to accessing, addressing or allocation within memory systems or architectures
- G06F2212/15—Use in a specific computing environment
- G06F2212/152—Virtualized environment, e.g. logically partitioned system
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2212/00—Indexing scheme relating to accessing, addressing or allocation within memory systems or architectures
- G06F2212/25—Using a specific main memory architecture
- G06F2212/254—Distributed memory
- G06F2212/2542—Non-uniform memory access [NUMA] architecture
Definitions
- Virtualization technology allows for sharing hardware resources between multiple partitions, each of which can host a guest operating system.
- virtual machine technology can be used to consolidate servers and increase their portability. As virtual machines become larger, and their workloads increase, the ability to easily consolidate and/or migrate them from one computer system to another becomes more difficult. Accordingly, techniques for increasing the ability to consolidate and/or migrate larger virtual machines are desirable.
- An example embodiment of the present disclosure describes a method.
- the method includes, but is not limited to receiving, by a management system, information indicative of the physical topology of computer systems in a datacenter; receiving a request to instantiate a virtual machine, the request indicating a characteristic for the virtual machine; determining, based on the physical topology of the computer systems and the characteristic for the virtual machine, a number of virtual NUMA nodes for the virtual machine; and sending a request to a computer system in the datacenter, the request directing the computer system to instantiate the virtual machine having the determined number of virtual NUMA nodes.
- An example embodiment of the present disclosure describes a method.
- the method includes, but is not limited to receiving a request to instantiate a virtual machine; generating the virtual machine, the virtual machine having a topology that includes a plurality of virtual NUMA nodes, each virtual NUMA node of the plurality including at least one virtual processor and a range of guest physical addresses, wherein the topology of the virtual machine is generated independently from the physical topology of the computer system; and instantiating the virtual machine, the virtual machine including the virtual NUMA nodes.
- a method includes, but is not limited to receiving a request to instantiate a virtual machine; generating the virtual machine, the virtual machine having a topology that includes a plurality of virtual NUMA nodes, each virtual NUMA node of the plurality including at least one virtual processor and a range of guest physical addresses, wherein the topology of the virtual machine is generated independently from the physical topology of the computer system; and instantiating the virtual machine, the virtual machine including the virtual NUMA nodes.
- the method includes, but is not limited to generating a plurality of virtual NUMA nodes for a virtual machine, each virtual NUMA node of the plurality including at least one virtual processor and at least one memory block of guest physical addresses, the plurality of virtual NUMA nodes being generating independently from a physical topology of any computer system; and instantiating the virtual machine on a computer system, the virtual machine including the plurality of virtual NUMA nodes.
- circuitry and/or programming for effecting the herein-referenced aspects of the present disclosure
- the circuitry and/or programming can be virtually any combination of hardware, software, and/or firmware configured to effect the herein-referenced aspects depending upon the design choices of the system designer.
- FIG. 1 depicts an example computer system wherein aspects of the present disclosure can be implemented.
- FIG. 2 depicts an operational environment for practicing aspects of the present disclosure.
- FIG. 3 depicts an operational environment for practicing aspects of the present disclosure.
- FIG. 4 depicts how memory can be arranged in embodiments.
- FIG. 5 depicts an example operational environment practicing aspects of the present disclosure.
- FIG.6 depicts an example operational environment practicing aspects of the present disclosure.
- FIG. 7 depicts an example operational environment practicing aspects of the present disclosure.
- FIG. 8 depicts an operational procedure for practicing aspects of the present disclosure.
- FIG. 9 depicts an alternative embodiment of the operational procedure 800 of FIG.
- FIG. 10 depicts an operational procedure for practicing aspects of the present disclosure.
- FIG. 11 depicts an alternative embodiment of the operational procedure 1000 of FIG. 10.
- FIG. 12 depicts an alternative embodiment of the operational procedure 1000 of FIG. 11.
- FIG. 13 depicts an operational procedure for practicing aspects of the present disclosure.
- FIG. 14 depicts an alternative embodiment of the operational procedure 1300 of FIG. 13.
- Embodiments may execute on one or more computers.
- FIG. 1 and the following discussion are intended to provide a brief general description of a suitable computing environment in which the disclosure may be implemented.
- the computer system of FIG. 1 can in some embodiments effectuate computer systems 200, 300, 600, and 700.
- the computer systems can include some or all of the components described in FIG. 1 and circuitry configured to instantiate aspects of the present disclosure.
- the term circuitry used through the disclosure can include hardware components such as hardware interrupt controllers, hard drives, network adaptors, graphics processors, hardware based video/audio codecs, and the firmware/software used to operate such hardware.
- circuitry can include microprocessors configured to perform function(s) by firmware or by switches set in a certain way.
- the term circuitry can include one or more logical processors, e.g., one or more cores of a multi-core general processing unit.
- the logical processor(s) in this example can be configured by software instructions embodying logic operable to perform function(s) that are loaded from memory, e.g., RAM, ROM, firmware, and/or virtual memory.
- an implementer may write source code embodying logic that is subsequently compiled into machine readable code that can be processed by a logical processor.
- the general purpose computing system can include a conventional computer 20 or the like, including a logical processor 21, a system memory 22, and a system bus 23 that couples various system components including the system memory to the logical processor 21.
- the system bus 23 may be any of several types of bus structures including a memory bus or memory controller, a peripheral bus, and a local bus using any of a variety of bus architectures.
- the system memory can include read only memory (ROM) 24 and random access memory (RAM) 25.
- ROM read only memory
- RAM random access memory
- the computer 20 may further include a hard disk drive 27 for reading from and writing to a hard disk (not shown), a magnetic disk drive 28 for reading from or writing to a removable magnetic disk 29, and an optical disk drive 30 for reading from or writing to a removable optical disk 31 such as a CD ROM or other optical media.
- the hard disk drive 27, magnetic disk drive 28, and optical disk drive 30 are shown as connected to the system bus 23 by a hard disk drive interface 32, a magnetic disk drive interface 33, and an optical drive interface 34, respectively.
- the drives and their associated computer readable storage media provide non volatile storage of computer readable instructions, data structures, program modules and other data for the computer 20.
- a number of program modules may be stored on the hard disk, magnetic disk 29, optical disk 31, ROM 24 or RAM 25, including an operating system 35, one or more application programs 36, other program modules 37 and program data 38.
- a user may enter commands and information into the computer 20 through input devices such as a keyboard 40 and pointing device 42.
- Other input devices may include a microphone, joystick, game pad, satellite disk, scanner or the like.
- serial port interface 46 that is coupled to the system bus, but may be connected by other interfaces, such as a parallel port, game port or universal serial bus (USB).
- a display 47 or other type of display device can also be connected to the system bus 23 via an interface, such as a video adapter 48.
- computers typically include other peripheral output devices (not shown), such as speakers and printers.
- the exemplary system of FIG. 1 also includes a host adapter 55, Small Computer System Interface (SCSI) bus 56, and an external storage device 62 connected to the SCSI bus 56.
- SCSI Small Computer System Interface
- the computer 20 may operate in a networked environment using logical connections to one or more remote computers, such as a remote computer 49.
- the remote computer 49 may be another computer, a server, a router, a network PC, a peer device or other common network node, and typically can include many or all of the elements described above relative to the computer 20, although only a memory storage device 50 has been illustrated in FIG. 1.
- the logical connections depicted in FIG. 1 can include a local area network (LAN) 51 and a wide area network (WAN) 52.
- LAN local area network
- WAN wide area network
- Such networking environments are commonplace in offices, enterprise wide computer networks, intranets and the Internet.
- the computer 20 When used in a LAN networking environment, the computer 20 can be connected to the LAN 51 through a network interface or adapter 53.
- the computer 20 can typically include a modem 54 or other means for establishing communications over the wide area network 52, such as the Internet.
- the modem 54 which may be internal or external, can be connected to the system bus 23 via the serial port interface 46.
- program modules depicted relative to the computer 20, or portions thereof may be stored in the remote memory storage device. It will be appreciated that the network connections shown are exemplary and other means of establishing a communications link between the computers may be used.
- numerous embodiments of the present disclosure are particularly well-suited for computerized systems, nothing in this document is intended to limit the disclosure to such embodiments. [0027] Referring now to FIG.
- computer system 200 can include physical hardware devices such as a storage device 208, e.g., a hard drive, a network interface controller (NIC) 210, a graphics card 234, at least one logical processor 212, and random access memory (RAM) 214.
- Computer system 200 can also include similar components as computer 20 of FIG. 1. While one logical processor is illustrated, in other embodiments computer system 200 may have multiple logical processors, e.g., multiple execution cores per processor and/or multiple processors that could each have multiple execution cores.
- a hypervisor 202 may also be referred to in the art as a virtual machine monitor.
- the hypervisor 202 in the depicted embodiment includes executable instructions for controlling and arbitrating access to the hardware of computer system 200.
- the hypervisor 202 can generate execution environments called partitions such as child partition 1 through child partition N (where N is an integer greater than 1).
- a child partition can be considered the basic unit of isolation supported by the hypervisor 202, that is, each child partition can be mapped to a set of hardware resources, e.g., memory, devices, logical processor cycles, etc., that is under control of the hypervisor 202 and/or the parent partition.
- the hypervisor 202 can be a stand-alone software product, a part of an operating system, embedded within firmware of the motherboard, specialized integrated circuits, or a combination thereof.
- the computer system 200 includes a parent partition 204 that can be also thought of as domain 0 in the open source community.
- Parent partition 204 can be configured to provide resources to guest operating systems executing in the child partitions 1-N by using virtualization service providers 228 (VSPs) that are also known as back-end drivers in the open source community.
- VSPs virtualization service providers
- the parent partition 204 can gate access to the underlying hardware.
- the VSPs 228 can be used to multiplex the interfaces to the hardware resources by way of virtualization service clients (VSCs) that are also known as front-end drivers in the open source community.
- VSCs virtualization service clients
- Each child partition can include one or more virtual processors such as virtual processors 230 through 232 that guest operating systems 220 through 222 can manage and schedule threads to execute thereon.
- the virtual processors 230 through 232 are executable instructions and associated state information that provide a representation of a physical processor with a specific architecture. For example, one virtual machine may have a virtual processor having characteristics of an Intel x86 processor, whereas another virtual processor may have the characteristics of a PowerPC processor.
- the virtual processors in this example can be mapped to logical processors of the computer system such that the instructions that effectuate the virtual processors will be backed by logical processors.
- multiple virtual processors can be simultaneously executing while, for example, another logical processor is executing hypervisor instructions.
- the combination of virtual processors, various VSCs, and memory in a partition can be considered a virtual machine such as virtual machine 240 or 242.
- guest operating systems 220 through 222 can include any operating system such as, for example, operating systems from Microsoft®, Apple®, the open source community, etc.
- the guest operating systems can include user/kernel modes of operation and can have kernels that can include schedulers, memory managers, etc.
- Each guest operating system 220 through 222 can have associated file systems that can have applications stored thereon such as e-commerce servers, email servers, etc., and the guest operating systems themselves.
- the guest operating systems 220-222 can schedule threads to execute on the virtual processors 230-232 and instances of such applications can be effectuated.
- FIG. 3 it illustrates an alternative architecture that can be used.
- FIG. 3 depicts similar components to those of FIG.
- hypervisor 202 can include the virtualization service providers 228 and device drivers 224, and parent partition 204 may contain configuration utilities 236.
- hypervisor 202 can perform the same or similar functions as the hypervisor 202 of FIG. 2.
- the hypervisor 202 of FIG. 3 can be a stand alone software product, a part of an operating system, embedded within firmware of the motherboard or a portion of hypervisor 202 can be effectuated by specialized integrated circuits.
- parent partition 204 may have instructions that can be used to configure hypervisor 202 however hardware access requests may be handled by hypervisor 202 instead of being passed to parent partition 204.
- FIG. 4 it illustrates how memory can be arranged in embodiments that include virtual machines.
- a computer system such as computer system 200 can have RAM 214 having memory addresses.
- hypervisor 202 can present different addresses for the system physical addresses, e.g., guest physical addresses
- GPAs to memory managers of guest operating systems.
- the guest operating systems can then manipulate the guest physical addresses and hypervisor 202 maintains the relationship by the GPAs and the SPAs.
- the GPAs and SPAs can be arranged into memory blocks.
- a memory block can include one or more pages of memory.
- the relationship between the GPAs and the SPAs can be maintained by shadow page table such those described in commonly assigned U.S. Patent Application No. 11/128,665 entitled "Enhanced Shadow Page Table Algorithms," the content of which is herein incorporated by reference by its entirety.
- shadow page table such those described in commonly assigned U.S. Patent Application No. 11/128,665 entitled "Enhanced Shadow Page Table Algorithms," the content of which is herein incorporated by reference by its entirety.
- the data may actually be stored in a different SPA such as block 6 on the system.
- FIG. 5 depicts an operational environment for practicing aspects of the present disclosure.
- a number of computer systems 504-510 can be coupled together in a datacenter 500 (although four computer systems are depicted, one of skill in the art can appreciate that the datacenter can include more or less computer system).
- the depicted computer systems can have different topologies and moreover, they can have different characteristics, e.g., different amounts of RAM, different RAM speeds, different amount of logical processors, and/or logical processors with different speeds.
- the management system 502 can have components similar to computer 20 of FIG. 1, and/or computer systems 200, 300, 600 or 700.. That is, in an embodiment the management system 502 may be a computer system that includes subject matter described below with respect to FIG. 6 or FIG. 7.
- FIG. 6 depicts a computer system 600 having a symmetric multiprocessing topology (SMP) or 'flat' topology.
- SMP is a computer architecture that includes a plurality of processors that are connected to a single shared memory.
- a memory controller 602 can manage the flow of data to and from memory. Memory access may be uniform with respect to each logical processor 212A-F and each logical processor can access the entire range of memory, i.e., system physical addresses 622-632.
- This topology works well for computer systems that have a relatively small number of processors, but when the computer system includes many processors, all competing for access to the shared memory bus, performance of the system can decrease.
- computer system 600 can include the same, or similar components as computer 200 or 300. As shown by the figure, computer system 600 can have a plurality of logical processors 212A-212F (while six logical processors are depicted the computer system could have more or less) coupled together via a memory controller 602 that gates access to RAM 214. Similar to that described above, each logical processor 212A-212F can have different characteristics, e.g., clock speeds, cache size, etc. In this arraignment the memory controller 602 can manage the flow of data to and from the RAM 214.
- Hypervisor 202 can be instantiated and it can control the hardware of the computer system 600.
- Hypervisor 202 can manage one or more virtual machines 240 through 242 that each can have virtual NUMA nodes such as virtual NUMA nodes 606-612.
- Virtual NUMA nodes 606-612 can be used to organize a virtual machine's resources by reporting virtual topologies to guest applications or guest operating systems such as guest operating systems 220 and 222.
- each virtual NUMA node 606-612 can have one or more virtual processors 230A-D, 232A-D and guest physical addresses 614- 616, and 618-620.
- hypervisor 202 can back each virtual NUMA node 606-612 with one or more logical processors and system physical addresses from RAM 214. That is, hypervisor 202 can set one or more logical processors as idea processors that can be used to run virtual processors threads.
- FIG. 7 depicts a computer system 700 having a topology that includes NUMA nodes 702-706.
- Computer systems with NUMA nodes are can be generally thought of as computers systems that are made up of smaller computer systems or cells.
- each NUMA node 606-612 can include one or more logical processors and local memory.
- NUMA node The memory inside of a NUMA node is considered local memory and memory in other NUMA nodes is considered remote memory because only the processors inside of the node are connected to the same memory bus.
- the NUMA nodes are interconnected by cache coherency domain interconnects which allow processors in one NUMA node to access memory in other NUMA nodes in a coherent way.
- system physical addresses 622-632 are uniform with respect to each processor.
- system physical address 20,000 is the same for every processor in the computer system. The difference is that for some processors memory address 20,000 is a local memory address, e.g., inside their NUMA node, and for other processors memory address 20,000 is remote, e.g., outside their NUMA node.
- Computer system 700 can include the same, or similar components as computer 200 or 300. As shown by the figure, in this operational environment computer system 700 includes three NUMA nodes 702-706 (although the computer can have more or less) connected by interconnects 708. Similar to that described above, the number of processors within each NUMA node can vary, the characteristics of the processors can vary, and each node can have its own RAM.
- hypervisor 202 can control the hardware of the computer system 700.
- guest operating systems or monolithic applications boot they can detect the topologies of the virtual machines 240 and 242 similar to that described above.
- Each virtual NUMA node 606-612 can be assigned one or more ideal processors and memory from one or more NUMA nodes.
- While computer system 600 and 700 are depicted as including two virtual machines 240 and 242, in other embodiments they can execute more or fewer virtual machines. Moreover, while each virtual machine is depicted has having two virtual NUMA nodes, in other embodiments the virtual machines can have more or fewer virtual NUMA nodes. Also, while virtual NUMA nodes are depicted as having two virtual processors, in other embodiments the virtual NUMA nodes can have more or fewer virtual processors. Furthermore, each virtual NUMA node can have a different topology than other virtual NUMA nodes, e.g., one virtual NUMA node can have 4 virtual processors and 8 gigabytes of RAM while another virtual NUMA node can have 2 virtual processors and 4 gigabytes of RAM.
- FIG. 8 it depicts an operational procedure for practicing aspects of the present disclosure including operations 800-808.
- the operational procedure begins with operation 800 and operation 802 illustrates receiving, by a management system, information indicative of the physical topology of computer systems in a datacenter. For example, and referring to FIG.
- a logical processor of management system 502 can execute one or more programs that can obtain information that identifies the physical topology of computer systems 504-510 in datacenter 500.
- management system 502 can include a NIC and the program can use the NIC to query each computer system 504-510.
- each computer system 504-510 can reply with information that identifies its topology.
- the information can be obtained from an administrator that inputs the information.
- the information may include, but is not limited to, information that identifies how many NUMA nodes (if any) each computer system has, how much RAM each computer system has, the speed of the RAM, how the RAM is arranged, processor speeds, how many cores each processor has, etc.
- operation 804 depicts receiving a request to instantiate a virtual machine, the request indicating a characteristic for the virtual machine.
- Management system 502 can receive a request to start virtual machine 240 from, for example, a user interface executing on management system 502 or from computer system 504-510.
- the request can include a desired characteristic for virtual machine 240.
- a characteristic could be a desired amount of RAM assigned to the virtual machine, a desired number of virtual processors, or desired I/O devices.
- operation 806 shows determining, based on the physical topology of the computer systems and the characteristic for the virtual machine, a number of virtual NUMA nodes for the virtual machine.
- the management system 502 can use this information along with the information that describes the topologies of the computer systems in the datacenter 500 to determine how many virtual NUMA nodes should be created for virtual machine 240. For example, management system 502 can determine the optimum size of a virtual NUMA node for datacenter 500 and then determine how many virtual NUMA nodes are needed to build a virtual machine having the desired characteristic.
- the size of the virtual NUMA node influences the operation of the virtual machine in datacenter 500. For example, as the size of a virtual NUMA node increases, e.g., in memory and/or processors, the portability of the virtual NUMA node decreases. Or put another way, large virtual NUMA nodes may make it more difficult to migrate the virtual machine. This occurs because a virtual NUMA node needs to be assigned to either a NUMA node or a computer system that has enough 'flat' resources to effectuate the virtual NUMA node.
- a virtual NUMA node is too large, e.g., it has too much RAM or too many virtual processors, it won't be able to fit in smaller NUMA nodes in the datacenter 500 thus limiting the ability to migrate the virtual machine.
- the larger virtual NUMA node is simply assigned to multiple smaller NUMA nodes performance of the virtual machine will decrease because of the difference that exists between local memory and remote memory access times and also because the guest operating system may not include information that indicates the difference in local v. remote memory access times.
- management system 502 can strike a balance between portability and efficiency by determining an optimum virtual NUMA node size for the datacenter 500.
- a logical processor of management system 502 can execute a program and determine the average size of a NUMA node in the datacenter, e.g., the average number of logical processors, average amount of RAM, etc., and set the size of the virtual NUMA node to be the same or smaller than the average NUMA node in the system.
- the program can be configured to set the size of the virtual NUMA node to be slightly smaller than the smallest NUMA node in the datacenter 500.
- the size of the virtual NUMA node can be set to be slightly smaller than the average size or smallest size so that if the computer system becomes heavily committed, more than one virtual NUMA node can be assigned to a single NUMA node. In a specific example, if the smallest NUMA node has 4 logical processors and 8 gigabytes of RAM, then the size of the virtual NUMA node could be set to, for example, 2 virtual processors and 4 gigabytes of RAM. [0048] Once the size of the virtual NUMA node is determined, and continuing from the example in the preceding paragraph, the size of the virtual NUMA node can be used along with the desired characteristic to generate virtual machine 240.
- management system 502 can generate a configuration file that indicates that the virtual machine will include 3 virtual NUMA nodes.
- operation 808 illustrates sending a request to a computer system in the datacenter, the request directing the computer system to instantiate the virtual machine having the determined number of virtual NUMA nodes.
- management system 502 can select a computer system in datacenter 500 to instantiate virtual machine 240 and send the configuration file to the selected computer system. For example, the management system 502 can determine which computer systems can accommodate the virtual machine, e.g., which computer has the physical resources to execute the virtual machine and which computer system has the lowest workload, and send virtual machine 240 to this computer.
- a user can indicate a priority for virtual machine 240, e.g., more critical virtual machines can be tagged with information that identifies their relative importance.
- management system 502 can be configured to send high priority virtual machines to computer systems with lower workloads and can attempt to consolidate lower priority virtual machines on a fewer computer systems.
- a virtual machine may be "bumped" off a computer system to make room for a higher priority virtual machine.
- management system 502 can receive a signal from a computer system such as, for example, computer system 504.
- the signal can indicate that the computer system's resources are stressed.
- the amount of RAM assigned to a virtual machine may be fluid. That is, it may be adjusted during runtime.
- the resources of computer system 504 may be overcommitted due to the fact that it is effectuating a large number of virtual machines, or virtual machines with heavy workloads.
- a hypervisor of computer system 504 may send a signal to the management system 502 that indicates that the computer system is low on memory, e.g., the signal may indicate that significant paging is occurring within virtual machines or that non-paged memory allocations may fail.
- guest operating system 220 and any workload in virtual machine 240 are taking an unacceptably long amount of time to execute.
- management system 502 can be configured to reduce the workload on the stressed computer system 504 by finding another computer that can accommodate virtual machine 240 and send a signal to the computer system 504 directing it to migrate virtual machine 240 to the new computer system, e.g., computer system 506.
- Hypervisor 202 can receive the signal and can migrate virtual machine 240 to a new computer system, e.g., to computer system 506, 508, or 510.
- operation 912 depicts sending a signal to the computer system, the signal directing the computer system to move the virtual machine from a first NUMA node to a second NUMA node.
- management system 502 can send a signal to computer system 504 directing it to move the virtual machine from one NUMA node to another.
- the architecture of computer system 504 may be similar to that of computer 700 of FIG. 7.
- Hypervisor 202 of FIG. 7 can receive a signal that directs it to move virtual machine 240 from one NUMA node to another NUMA node, e.g., from NUMA node 704 to 702.
- memory pressure in guest OS 220 may be high or hypervisor threads for virtual processors 23 OA-D may not be scheduled fast enough on logical processors 212E and 212G.
- management system 502 can scan datacenter 500 and determine that virtual machine 240 should be moved to NUMA node 702 for example and hypervisor 202 can reassign virtual NUMA nodes 606 and 608 to NUMA node 702.
- operation 806 can include operation 914 which depicts determining the highest NUMA ratio in the datacenter.
- a processor of management system 502 can execute a program and determine the highest NUMA ratio that exists in datacenter 500 and use the information when determining the number of virtual NUMA nodes for virtual machine 240.
- the NUMA ratio is a cost ratio in processor cycles between accessing remote vs. local memory.
- a NUMA ratio of 2 to 1 means that it takes twice as many processor cycles to access a specific remote memory address than it costs to access a local memory address.
- management system 502 can use the highest detected NUMA ratio in datacenter 500 to set the NUMA ratio reported by virtual NUMA nodes 606-608. This information can be stored in the configuration file and sent to the computer system that is going to instantiate the virtual machine.
- Hypervisor 202 can create a data structure in the virtual machine's firmware table that indicates the NUMA ratio for virtual NUMA nodes 606-608 and when guest operating system 220, or monolithic application boots the NUMA ratio can be used to make thread scheduling decisions.
- Operation 1000 begins the operational procedure and operation 1002 illustrates receiving a request to instantiate a virtual machine.
- hypervisor 202 can receive a request to create a virtual machine such as virtual machine 240.
- the request can be received from a management system 502, a parent partition 204 of FIG. 2 or 3, etc.
- the request can be for a new virtual machine or it can be a request to instantiate a previously saved virtual machine.
- the characteristics of virtual machine 240 can be set by, for example, an administrator.
- operation 1004 it depicts generating the virtual machine, the virtual machine having a topology that includes a plurality of virtual NUMA nodes, each virtual NUMA node of the plurality including at least one virtual processor and a range of guest physical addresses, wherein the topology of the virtual machine is generated independently from the physical topology of the computer system.
- hypervisor 202 can construct virtual machine 240 having a plurality of virtual NUMA nodes 606-608 in response to the request.
- virtual NUMA nodes 606- 608 can each have one or more virtual processors 230A-D and guest physical addresses 614 and 616.
- virtual NUMA nodes 606 and 608 can be created independently from the topology of the underlying hardware. That is, the topology of the virtual machine is unrelated to the underlying hardware such as is depicted by FIG. 6 and FIG. 7. Thus, in this embodiment, each virtual machine's topology is decoupled from the underlying physical topology of the computer system that effectuates it.
- operation 1006 illustrates instantiating the virtual machine, the virtual machine including the virtual NUMA nodes.
- hypervisor 202 can be executed by a logical processor and a virtual machine having the plurality of virtual NUMA nodes can be instantiated.
- a virtual machine 240 having virtual NUMA nodes 606- 608 can be effectuated by computer system 600 or 700.
- Guest physical addresses of the virtual machine 240 can be backed with system physical addresses from RAM and virtual processors can be backed by one or more logical processors.
- blocks of guest physical addresses 614 can be backed with blocks of system physical addresses 622 and blocks of guest physical addresses 616 can be backed by blocks of system physical addressees 624.
- Hypervisor threads can then be scheduled on the logical processors backing the virtual processors and instructions indicative of the virtual processors can be executed.
- guest operating system 220 can schedule a process to execute on virtual processor 230A of virtual NUMA node 606.
- the memory manager of guest operating system 220 can select a range of guest physical addresses and allocate the range of addresses to the process.
- guest operating system 220 can be configured to select guest physical addresses from virtual NUMA node 606. In this case the ability to migrate/restore the virtual machine is enhanced due to the fact that the virtual machine is broken up into smaller pieces and spread across the computer system.
- the virtual machine BIOS or boot firmware can describe the virtual machine's topology, e.g., whether it has virtual NUMA nodes, the size of any virtual NUMA nodes, and NUMA ratios for the virtual NUMA nodes, to a guest operating system of monolithic application.
- the data structure can be processed and the guest OS or application can configure itself to take advantage of the presence of virtual NUMA nodes.
- guest operating system 220 can try to aff ⁇ nitize the threads of an application that is not NUMA aware to a virtual NUMA node so that execution of the application remains local.
- a database management program such as SQL Server can allocate locks local to the virtual NUMA nodes and the database can split up read/write requests across the virtual NUMA nodes.
- the guest operating system 220 can create page pools for each virtual NUMA node in the virtual machine.
- operation 1108 illustrates saving the virtual machine.
- virtual machine 240 can be saved to a computer readable storage medium such as a hard drive.
- the virtual machine 240 can be saved using techniques described in U.S. Patent Application No. 11/487,031 entitled “Concurrent Virtual Machine Snapshots and Restore,” the content of which is herein incorporated by reference in its entirety.
- operation 1110 shows sending the virtual machine to a remote computer system. For example, and referring to FIG.
- the state of virtual machine 240 can be saved in one or more configuration files and sent to another computer system, e.g., sent from computer system 504 to 506.
- the hypervisor of computer system 506 can read the file or files and instantiate virtual machine 240.
- hypervisor of computer system 504 can determine to send the virtual machine 240 to computer system 506.
- Virtual machine topology affects the ability to migrate and restore virtual machines. Specifically, the decision to allow the topology of the underlying hardware to be detected and the size of the virtual NUMA nodes influence how well a virtual machine will execute and whether it can be easily migrated. For example, the size of the virtual NUMA nodes affects the ability to migrate the virtual machine. That is, as the size of a virtual NUMA node increases the portability of the virtual NUMA node decrease and as the size of the virtual NUMA node decreases so does performance of the virtual machine.
- virtual machines that can detect the underlying computer's topology can not be easily migrated due to the fact that NUMA aware operating systems and applications optimize themselves at boot time based on first topology they detect and these optimizations may not work well on computers that the virtual machine may be migrated to in the future.
- virtual NUMA nodes to the guest operating system when it boots the operating system can be optimized to use NUMA nodes.
- the virtual machine can be optimized for many diverse computer systems in datacenter 500.
- virtual machine 240 can include two or more virtual NUMA nodes 606 and 608.
- Hypervisor 202 can back the virtual processors 230A- D with logical processors 212A-D.
- guest operating system 220 boots it can detect virtual NUMA nodes 606 and 608 and can be configured to optimize the scheduling and execution of processes.
- virtual machine 240 may be migrated to a computer system having a physical topology similar to that depicted by FIG. 7.
- Hypervisor 202 of FIG. 7 can back the virtual processors 230A and B with logical processors 212A and B and back virtual processors 230C and D with logical processors 212E and F.
- Guest operating system 220 can continue to operate in the same manner as it was when running on the computer system of FIG. 6 even though the underlying computer topology has changed from SMP to NUMA.
- operation 1112 shows migrating the virtual machine to one or more other NUMA nodes.
- hypervisor 202 can schedule virtual NUMA nodes 606-608 to run on NUMA node 702 and sometime later schedule virtual NUMA nodes 606-608 to run on, for example, NUMA node 704.
- hypervisor 202 may migrate virtual machine 240 when NUMA node 702 is stressed.
- guest operating system 220 and 222 may generate signals that indicate that virtual machine 240 is low on memory.
- hypervisor 202 can be configured to reduce the workload on NUMA node 702 by moving virtual machine 240 to a different NUMA node.
- operation 1114 shows assigning the virtual machine to a first NUMA node; and migrating the virtual machine to a second
- NUMA node of the plurality of NUMA nodes can be assigned to first NUMA node 606 by hypervisor 202. That is, hypervisor instructions can be executed by a logical processor 212A - 2121 and virtual machine 240 can be assigned to, for example, NUMA node 702.
- virtual processors 230A-230D may be set to execute on logical processors 212A through 212D.
- hypervisor 202 can be executed by a logical processor 212A - 2121 and the logical processor can migrate virtual machine 240 to another NUMA node in the computer system 700.
- hypervisor 202 can be executed and virtual machine 240 can be moved from NUMA node 702 to NUMA node 704.
- virtual processor 230A and B may be assigned to logical processor 212E
- virtual processor 230C and D may be assigned to logical processor 212F
- guest physical addresses 614 and 616 can be backed by system physical addresses 622-624.
- operation 1116 shows generating NUMA ratios for the plurality of virtual NUMA nodes; and reporting the NUMA ratios for the plurality of virtual NUMA nodes to a guest operating system.
- hypervisor 202 can generate NUMA ratios for virtual NUMA nodes 606-608 and this information can be reported to guest operating system 220 of either FIG. 6 or FIG. 7.
- hypervisor 202 can create a data structure in a virtual machine's firmware table that indicates the NUMA ratio for the virtual NUMA nodes and when the guest operating system 220 boots the guest can read the table and use the information to make thread scheduling decisions.
- a guest operating system, or NUMA aware application may use the NUMA ratios to determine whether or not to use resources from a remote NUMA node. For example, an operating system may have a pending thread that is ready to be executed.
- the operating system in this example can also be configured to wait a certain amount of time for an ideal processor to become free otherwise it will schedule the thread on a remote processor that has less than a predetermined NUMA ratio. In this case the amount of time the scheduler is willing to wait is dependent upon the NUMA ratio.
- operation 1118 shows receiving a request from a guest operating system to access a first guest physical address from a first virtual NUMA node and a second guest physical address from a second virtual NUMA node; increasing the access time required to access the second guest physical address; and reporting the increased access time required to access the first and second guest physical addresses to the guest operating system.
- NUMA ratios can be reported to guest operating system 220, however in this example embodiment guest operating system 220 may include instructions for determining whether the NUMA ratios are accurate.
- hypervisor 202 can be configured to monitor the boot process of virtual machine 240 and hypervisor 202 can slow down access to memory that is remote to a particular virtual NUMA node. That is, hypervisor 202 can be configured to generate larger access times for when virtual processor 230A accesses guest physical addresses 616 than when virtual processor 230A accesses guest physical addresses 614. In this example, the delay may be fictional because guest physical addresses 614 and 616 may be backed by system physical addresses from a single NUMA node such as system physical addresses 622 and 624.
- hypervisor 202 can increase the access time by removing remote guest physical addressees 616 from the virtual processor page tables of virtual processor 230A. When virtual processor 230A attempts to access the remote memory it would fault and hypervisor instructions can be executed. Hypervisor 202 can then fix the page tables and delay the return signal from the memory address for a desired amount of time. When virtual processor 230A receives the return signal information that identifies how long it took to access the remote memory delay is introduced.
- operation 1120 depicts allocating system memory addresses to the virtual machine, the system memory addresses obtained from a single NUMA node; and assigning a processor to execute the virtual processor, the processor obtained from the NUMA node.
- generating virtual NUMA nodes 606-608 can include allocating system physical addresses and assigning a processor from the same NUMA node 702 to back the virtual machine 240.
- the hypervisor 202 can obtain one or more blocks of system physical addresses 622-624 of NUMA node 702 and use them to back guest physical addresses 622-624.
- Hypervisor 202 can additionally aff ⁇ nitize virtual processors 230A to, for example, one or more logical processors 212A-212D of NUMA node 702.
- operation 1122 depicts mapping a first virtual NUMA node of the plurality to a first NUMA node of the computer system; and mapping a second virtual NUMA node of the plurality to the first NUMA node of the computer system.
- a logical processor such as logical processor 212A can execute hypervisor instructions and can affmitize virtual NUMA nodes such as virtual NUMA node 606 and 608 to a NUMA node such as NUMA node 702.
- the logical processor execute instructions and can back guest physical addresses 614-616 with system physical addresses 622, guest physical addresses 618-620 with system physical addresses 624 and back virtual processors 230A through D and 232A-D with logical processors 212A through D.
- mapping operation 1124 it depicts mapping a first virtual NUMA node of the plurality to a first NUMA node of the computer system; and mapping a second virtual NUMA node of the plurality to a second NUMA node of the computer system.
- a logical processor such as logical processor 212A can execute hypervisor instructions and can assign a virtual NUMA node 606 to NUMA node 702 and assign virtual NUMA node 608 to NUMA node 704.
- a hypervisor scheduler can schedule threads from virtual processors 23 OA-B on logical processors 212A-D and schedule threads from virtual processor 230C or D on logical processors 212E through G.
- the procedure can include operation 1226 which shows that in an embodiment the NUMA ratios for the virtual NUMA nodes are greater than a median NUMA ratio detected in a datacenter, the datacenter including a plurality of computer systems.
- a management system 502 can obtain the physical topology of the computer systems 504-510 in datacenter 500.
- the program can query each computer system 504-510 and the computers can reply with the information.
- the program can determine which computer system has a NUMA node with the highest, e.g., the worst, NUMA ratio, and/or the median NUMA ratio based.
- the management system 502 can use this information to generate the NUMA ratio that the virtual NUMA nodes will report to guest operating systems at boot time.
- the management system 502 can set the virtual NUMA ratio to be the median or greater than the highest NUMA ratio detected in the datacenter 500 thus, configuring the virtual NUMA nodes to reflect a state that is higher than half of the NUMA nodes or worse than any physical topology that any virtual machine will actually run on.
- Operation 1300 begins the operational procedure and operation 1302 depicts generating a plurality of virtual NUMA nodes for a virtual machine, each virtual NUMA node of the plurality including at least one virtual processor and at least one memory block of guest physical addresses, the plurality of virtual NUMA nodes being generating independently from a physical topology of any computer system.
- a logical processor such as 212E of FIG. 6 or FIG. 7 can execute hypervisor instructions and the processor can generate virtual NUMA nodes 606-608 for virtual machine 240.
- virtual machine 240 may have many more virtual NUMA nodes depending on, for example, the characteristics of the virtual machine, e.g., the virtual machine may have two or more virtual NUMA nodes.
- Hypervisor 202 can allocate the guest physical addresses to each virtual NUMA node 606 and 608.
- the topology of virtual machine 240 is independent from the underlying topology of the computer system.
- the computer system that the virtual machine is eventually instantiated on can have a topology similar to computer system 600 of FIG. 6 or computer system 700 of FIG. 7.
- operation 1304 shows instantiating the virtual machine on a computer system, the virtual machine including the plurality of virtual NUMA nodes.
- hypervisor 202 schedule instructions for starting the virtual machine 240.
- hypervisor 202 can select one or more logical processors to run instructions from virtual machine 240 and the hypervisor 202 can schedule instructions for instantiating virtual machine 240 on a logical processor.
- Virtual machine instructions can be executed by a processor and virtual machine 240 can be instantiated.
- Virtual machine 240 in this example can include boot tables that describe two or more virtual NUMA nodes.
- a guest operating system, or monolithic application can boot and discover the virtual NUMA nodes from the boot tables and execute instructions to configure themselves to take advantage of the topology of the virtual machine 240.
- FIG. 14 it depicts an alternative embodiment of the operational procedure 1300 of FIG. 13 including the additional operations 1406-1412.
- operation 1406 it illustrates assigning a first virtual NUMA node and a second virtual NUMA node to execute on a NUMA node of the computer system.
- computer system can have an architecture that includes NUMA nodes.
- hypervisor 202 can assign two virtual NUMA nodes of the plurality to a signal NUMA node.
- FIG. 14 depicts an alternative embodiment of the operational procedure 1300 of FIG. 13 including the additional operations 1406-1412.
- operation 1406 it illustrates assigning a first virtual NUMA node and a second virtual NUMA node to execute on a NUMA node of the computer system.
- computer system can have an architecture that includes NUMA nodes.
- hypervisor 202 can assign two virtual NUMA nodes
- hypervisor instructions can be executed by a logical processor and virtual NUMA node 606 and 608 can be assigned to run on NUMA node 702 for example.
- operation 1408 depicts assigning a first virtual NUMA node of the plurality to execute on a first NUMA node of the computer system; and assigning a second virtual NUMA node of the plurality to execute on a second NUMA node of the computer system.
- a logical processor such as logical processor 212B can execute hypervisor instructions and can assign a virtual NUMA node 606 to NUMA node 704 and assign virtual NUMA node 608 to NUMA node 706.
- a hypervisor scheduler can schedule threads from virtual processors 23 OA-B on logical processors 212F and 212G and schedule threads from virtual processor 230C or D on logical processors 212H and 2121.
- operation 1410 depicts generating
- NUMA ratios for the plurality of virtual NUMA nodes and reporting the NUMA ratios to a guest operating system.
- the hypervisor 202 can generate NUMA ratios for the virtual NUMA nodes and this information can be reported to the guest operating system 220 of either FIG. 6 or FIG. 7.
- operation 1412 depicts determining that a NUMA node associated with the virtual machine is overcommitted; and migrating the virtual machine to one or more other NUMA nodes.
- hypervisor 202 can schedule virtual NUMA nodes 606-612 to run on NUMA node 702. That is, NUMA node 702 may be configured to host virtual machines 240 and 242.
- hypervisor 202 can receive a signal that indicates that NUMA node 702 is stressed. For example, guest operating system 220 and 222 may generate signals that indicate that virtual machines 240-242 are low on memory. In this example, hypervisor 202 can be configured to reduce the workload on NUMA node 702 by moving a virtual machine to a different NUMA node. In this example, hypervisor 202 can reassign the virtual NUMA nodes 606 and 608 of virtual machine 240 to execute on NUMA node 704. [0076] The foregoing detailed description has set forth various embodiments of the systems and/or processes via examples and/or operational diagrams.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Human Computer Interaction (AREA)
- Stored Programmes (AREA)
- Memory System Of A Hierarchy Structure (AREA)
Abstract
Description
Claims
Priority Applications (7)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP10756651.5A EP2411916B1 (en) | 2009-03-26 | 2010-03-19 | Virtual non-uniform memory architecture for virtual machines |
KR1020117022331A KR101722276B1 (en) | 2009-03-26 | 2010-03-19 | Virtual non-uniform memory architecture for virtual machines |
JP2012502135A JP5507661B2 (en) | 2009-03-26 | 2010-03-19 | Non-uniform virtual memory architecture for virtual machines |
CA2753229A CA2753229C (en) | 2009-03-26 | 2010-03-19 | Virtual non-uniform memory architecture for virtual machines |
RU2011139102/08A RU2571366C2 (en) | 2009-03-26 | 2010-03-19 | Virtual non-uniform memory access architecture for virtual machines |
CN2010800137822A CN102365626B (en) | 2009-03-26 | 2010-03-19 | Virtual non-uniform memory architecture for virtual machines |
BRPI1009274A BRPI1009274A2 (en) | 2009-03-26 | 2010-03-19 | Uniformity virtual memory architecture for virtual machines |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US12/412,258 | 2009-03-26 | ||
US12/412,258 US9535767B2 (en) | 2009-03-26 | 2009-03-26 | Instantiating a virtual machine with a virtual non-uniform memory architecture |
Publications (2)
Publication Number | Publication Date |
---|---|
WO2010111151A2 true WO2010111151A2 (en) | 2010-09-30 |
WO2010111151A3 WO2010111151A3 (en) | 2011-01-13 |
Family
ID=42781792
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/US2010/028038 WO2010111151A2 (en) | 2009-03-26 | 2010-03-19 | Virtual non-uniform memory architecture for virtual machines |
Country Status (9)
Country | Link |
---|---|
US (2) | US9535767B2 (en) |
EP (1) | EP2411916B1 (en) |
JP (1) | JP5507661B2 (en) |
KR (1) | KR101722276B1 (en) |
CN (1) | CN102365626B (en) |
BR (1) | BRPI1009274A2 (en) |
CA (2) | CA3003789C (en) |
RU (1) | RU2571366C2 (en) |
WO (1) | WO2010111151A2 (en) |
Families Citing this family (45)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9535767B2 (en) * | 2009-03-26 | 2017-01-03 | Microsoft Technology Licensing, Llc | Instantiating a virtual machine with a virtual non-uniform memory architecture |
US9529636B2 (en) | 2009-03-26 | 2016-12-27 | Microsoft Technology Licensing, Llc | System and method for adjusting guest memory allocation based on memory pressure in virtual NUMA nodes of a virtual machine |
US8780923B2 (en) * | 2010-01-15 | 2014-07-15 | Dell Products L.P. | Information handling system data center bridging features with defined application environments |
US8904382B2 (en) * | 2010-06-17 | 2014-12-02 | International Business Machines Corporation | Creating instances of cloud computing environments |
US9733963B2 (en) * | 2010-09-17 | 2017-08-15 | Lenovo Enterprise Solutions (Singapore) Pte. Ltd. | Optimizing virtual graphics processing unit utilization |
US8667441B2 (en) | 2010-11-16 | 2014-03-04 | International Business Machines Corporation | Clock optimization with local clock buffer control optimization |
KR101145144B1 (en) * | 2010-11-23 | 2012-05-14 | 한국과학기술정보연구원 | Method and system for scheduling virtual machine |
US9250863B1 (en) * | 2010-12-28 | 2016-02-02 | Amazon Technologies, Inc. | Managing virtual machine migration |
US8725483B2 (en) | 2011-01-19 | 2014-05-13 | International Business Machines Corporation | Minimizing the maximum required link capacity for three-dimensional interconnect routing |
US8621057B2 (en) * | 2011-03-07 | 2013-12-31 | International Business Machines Corporation | Establishing relationships among elements in a computing system |
US9183030B2 (en) * | 2011-04-27 | 2015-11-10 | Microsoft Technology Licensing, Llc | Virtual processor allocation techniques |
US20130007370A1 (en) * | 2011-07-01 | 2013-01-03 | Oracle International Corporation | Method and apparatus for minimizing working memory contentions in computing systems |
US8856495B2 (en) * | 2011-07-25 | 2014-10-07 | International Business Machines Corporation | Automatically routing super-compute interconnects |
US9164809B2 (en) * | 2012-09-04 | 2015-10-20 | Red Hat Israel, Ltd. | Virtual processor provisioning in virtualized computer systems |
US20140229940A1 (en) * | 2013-02-14 | 2014-08-14 | General Dynamics C4 Systems, Inc. | Methods and apparatus for synchronizing multiple processors of a virtual machine |
US10061622B2 (en) * | 2013-02-26 | 2018-08-28 | Red Hat Israel, Ltd. | Updating memory topology information for virtual machines |
US10114662B2 (en) * | 2013-02-26 | 2018-10-30 | Red Hat Israel, Ltd. | Updating processor topology information for virtual machines |
US9348654B2 (en) * | 2013-11-19 | 2016-05-24 | International Business Machines Corporation | Management of virtual machine migration in an operating environment |
JP6287321B2 (en) * | 2014-02-24 | 2018-03-07 | 富士通株式会社 | Information processing apparatus, information processing method, and information processing program |
US9417902B1 (en) * | 2014-03-24 | 2016-08-16 | Amazon Technologies, Inc. | Managing resource bursting |
US9223574B2 (en) * | 2014-03-27 | 2015-12-29 | International Business Machines Corporation | Start virtual execution instruction for dispatching multiple threads in a computer |
US10146634B1 (en) * | 2014-03-31 | 2018-12-04 | EMC IP Holding Company LLC | Image restore from incremental backup |
WO2015193947A1 (en) * | 2014-06-16 | 2015-12-23 | 株式会社日立製作所 | Physical computer and virtual computer transfer method |
US10079797B2 (en) * | 2014-10-29 | 2018-09-18 | Vmware, Inc. | Methods, systems and apparatus to remotely start a virtual machine |
CN105677454B (en) * | 2014-11-20 | 2019-08-27 | 华为技术有限公司 | The integration method of computing resource, device and system |
US10140218B2 (en) * | 2015-01-15 | 2018-11-27 | Red Hat Israel, Ltd. | Non-uniform memory access support in a virtual environment |
US9959233B2 (en) * | 2015-04-28 | 2018-05-01 | Red Hat Israel, Ltd. | NUMA-aware root bus selection |
US9798497B1 (en) * | 2015-06-08 | 2017-10-24 | Skytap | Storage area network emulation |
US10241674B2 (en) * | 2015-12-11 | 2019-03-26 | Vmware, Inc. | Workload aware NUMA scheduling |
JP6645275B2 (en) * | 2016-03-04 | 2020-02-14 | 富士通株式会社 | Information processing apparatus, virtual machine moving method, and virtual machine control program |
US10095550B2 (en) * | 2016-10-19 | 2018-10-09 | International Business Machines Corporation | Performance-based reallocating of logical processing units to sockets of a computer system |
CN114218012A (en) * | 2017-04-17 | 2022-03-22 | 伊姆西Ip控股有限责任公司 | Method and system for backing up and restoring data in cluster system |
CN107168771A (en) * | 2017-04-24 | 2017-09-15 | 上海交通大学 | A kind of scheduling virtual machine device and method under Non Uniform Memory Access access architectures |
EP3411785B1 (en) * | 2017-04-28 | 2020-09-02 | LZLabs GmbH | Containerized deployment of microservices based on monolithic legacy applications |
US10649943B2 (en) | 2017-05-26 | 2020-05-12 | Dell Products, L.P. | System and method for I/O aware processor configuration |
US10613990B2 (en) * | 2017-12-05 | 2020-04-07 | Red Hat, Inc. | Host address space identifier for non-uniform memory access locality in virtual machines |
US10430227B2 (en) * | 2018-01-09 | 2019-10-01 | Microsoft Technology Licensing, Llc | Fast instantiation of virtual machines in distributed computing systems |
US10664323B2 (en) | 2018-03-07 | 2020-05-26 | Microsoft Technology Licensing, Llc | Live migration of virtual machines in distributed computing systems |
US10901911B2 (en) | 2018-11-21 | 2021-01-26 | Microsoft Technology Licensing, Llc | Faster computer memory access by reducing SLAT fragmentation |
US10503543B1 (en) * | 2019-02-04 | 2019-12-10 | Cohesity, Inc. | Hosting virtual machines on a secondary storage system |
US12020052B2 (en) * | 2019-10-30 | 2024-06-25 | Hewlett Packard Enterprise Development Lp | Goal-directed software-defined NUMA working set management |
CN114090223A (en) * | 2020-08-24 | 2022-02-25 | 北京百度网讯科技有限公司 | Memory access request scheduling method, device, equipment and storage medium |
US11740921B2 (en) * | 2020-11-23 | 2023-08-29 | Google Llc | Coordinated container scheduling for improved resource allocation in virtual computing environment |
US11928502B2 (en) * | 2021-04-29 | 2024-03-12 | VMware LLC | Optimized networking thread assignment |
KR102676385B1 (en) * | 2021-11-01 | 2024-06-19 | 한화시스템 주식회사 | Apparatus and method for managing virtual machine cpu resource in virtualization server |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20050120160A1 (en) | 2003-08-20 | 2005-06-02 | Jerry Plouffe | System and method for managing virtual servers |
Family Cites Families (44)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH04274519A (en) * | 1991-02-28 | 1992-09-30 | Nec Corp | Automatically executing system for accelerating program |
US6075938A (en) | 1997-06-10 | 2000-06-13 | The Board Of Trustees Of The Leland Stanford Junior University | Virtual machine monitors for scalable multiprocessors |
US6785888B1 (en) | 1997-08-29 | 2004-08-31 | International Business Machines Corporation | Memory allocator for a multiprocessor computer system |
US6336170B1 (en) * | 1998-10-13 | 2002-01-01 | International Business Machines Corporation | Method and system in a distributed shared-memory data processing system for determining utilization of shared-memory included within nodes by a designated application |
US7140020B2 (en) | 2000-01-28 | 2006-11-21 | Hewlett-Packard Development Company, L.P. | Dynamic management of virtual partition computer workloads through service level optimization |
US7433951B1 (en) | 2000-09-22 | 2008-10-07 | Vmware, Inc. | System and method for controlling resource revocation in a multi-guest computer system |
US7035963B2 (en) | 2000-12-27 | 2006-04-25 | Intel Corporation | Method for resolving address space conflicts between a virtual machine monitor and a guest operating system |
JP2002202959A (en) | 2000-12-28 | 2002-07-19 | Hitachi Ltd | Virtual computer system for performing dynamic resource distribution |
US6785773B2 (en) | 2001-03-29 | 2004-08-31 | International Business Machines Corporation | Verification of global coherence in a multi-node NUMA system |
US6842857B2 (en) | 2001-04-12 | 2005-01-11 | International Business Machines Corporation | Method and apparatus to concurrently boot multiple processors in a non-uniform-memory-access machine |
US7536697B2 (en) | 2001-06-19 | 2009-05-19 | Accenture Global Services Gmbh | Integrating enterprise support systems |
US6886079B2 (en) | 2001-06-21 | 2005-04-26 | International Business Machines Corporation | Dynamic history based mechanism for the granting of exclusive data ownership in a non-uniform memory access (NUMA) computer system |
US7159216B2 (en) * | 2001-11-07 | 2007-01-02 | International Business Machines Corporation | Method and apparatus for dispatching tasks in a non-uniform memory access (NUMA) computer system |
DE10159522A1 (en) | 2001-12-05 | 2003-06-26 | G L I Global Light Ind Gmbh | Light conducting LED members are formed by locating electrical components in an injection mould, evacuating the mould, and injecting a material through an injection opening |
US7451459B2 (en) | 2003-05-05 | 2008-11-11 | Microsoft Corporation | Systems, methods, and apparatus for indicating processor hierarchical topology |
US8417913B2 (en) | 2003-11-13 | 2013-04-09 | International Business Machines Corporation | Superpage coalescing which supports read/write access to a new virtual superpage mapping during copying of physical pages |
US20050132362A1 (en) | 2003-12-10 | 2005-06-16 | Knauerhase Robert C. | Virtual machine management using activity information |
US7231504B2 (en) | 2004-05-13 | 2007-06-12 | International Business Machines Corporation | Dynamic memory management of unallocated memory in a logical partitioned data processing system |
WO2005116828A2 (en) | 2004-05-21 | 2005-12-08 | Computer Associates Think, Inc. | Method and apparatus for dynamic memory resource management |
US8621458B2 (en) * | 2004-12-21 | 2013-12-31 | Microsoft Corporation | Systems and methods for exposing processor topology for virtual machines |
US20060184938A1 (en) | 2005-02-17 | 2006-08-17 | Intel Corporation | Method, apparatus and system for dynamically reassigning memory from one virtual machine to another |
US7334076B2 (en) | 2005-03-08 | 2008-02-19 | Microsoft Corporation | Method and system for a guest physical address virtualization in a virtual machine environment |
US20060206891A1 (en) | 2005-03-10 | 2006-09-14 | International Business Machines Corporation | System and method of maintaining strict hardware affinity in a virtualized logical partitioned (LPAR) multiprocessor system while allowing one processor to donate excess processor cycles to other partitions when warranted |
RU47116U1 (en) | 2005-04-06 | 2005-08-10 | Оао Акб "Автобанк-Никойл" | DISTRIBUTED DOCUMENT CIRCUIT SUPPORT SYSTEM |
US7299337B2 (en) | 2005-05-12 | 2007-11-20 | Traut Eric P | Enhanced shadow page table algorithms |
US20070100845A1 (en) | 2005-10-31 | 2007-05-03 | Juergen Sattler | Customer relationship management integration system and method |
US7945913B2 (en) | 2006-01-19 | 2011-05-17 | International Business Machines Corporation | Method, system and computer program product for optimizing allocation of resources on partitions of a data processing system |
US7596654B1 (en) * | 2006-01-26 | 2009-09-29 | Symantec Operating Corporation | Virtual machine spanning multiple computers |
US7984438B2 (en) | 2006-02-08 | 2011-07-19 | Microsoft Corporation | Virtual machine transitioning from emulating mode to enlightened mode |
US20070250784A1 (en) | 2006-03-14 | 2007-10-25 | Workstone Llc | Methods and apparatus to combine data from multiple computer systems for display in a computerized organizer |
JP4702127B2 (en) | 2006-03-22 | 2011-06-15 | 日本電気株式会社 | Virtual computer system, physical resource reconfiguration method and program thereof |
US7434002B1 (en) | 2006-04-24 | 2008-10-07 | Vmware, Inc. | Utilizing cache information to manage memory access and cache utilization |
US8607009B2 (en) | 2006-07-13 | 2013-12-10 | Microsoft Corporation | Concurrent virtual machine snapshots and restore |
US8453132B2 (en) | 2006-07-28 | 2013-05-28 | Hewlett-Packard Development Company, L.P. | System and method for recompiling code based on locality domain and thread affinity in NUMA computer systems |
US20080071642A1 (en) | 2006-09-15 | 2008-03-20 | Leiba Lior | System and method for connecting external product catalog data to business applications |
US7788464B2 (en) | 2006-12-22 | 2010-08-31 | Microsoft Corporation | Scalability of virtual TLBs for multi-processor virtual machines |
US8195866B2 (en) | 2007-04-26 | 2012-06-05 | Vmware, Inc. | Adjusting available persistent storage during execution in a virtual computer system |
CN100489815C (en) | 2007-10-25 | 2009-05-20 | 中国科学院计算技术研究所 | EMS memory sharing system, device and method |
CN101271401B (en) | 2008-04-23 | 2010-04-14 | 北京航空航天大学 | Server cluster unit system with single system image |
US8799892B2 (en) | 2008-06-09 | 2014-08-05 | International Business Machines Corporation | Selective memory donation in virtual real memory environment |
US8037280B2 (en) * | 2008-06-11 | 2011-10-11 | Vmware, Inc. | System and method for improving memory locality of virtual machines |
US8490094B2 (en) * | 2009-02-27 | 2013-07-16 | International Business Machines Corporation | Non-uniform memory access (NUMA) enhancements for shared logical partitions |
US9535767B2 (en) | 2009-03-26 | 2017-01-03 | Microsoft Technology Licensing, Llc | Instantiating a virtual machine with a virtual non-uniform memory architecture |
US9529636B2 (en) | 2009-03-26 | 2016-12-27 | Microsoft Technology Licensing, Llc | System and method for adjusting guest memory allocation based on memory pressure in virtual NUMA nodes of a virtual machine |
-
2009
- 2009-03-26 US US12/412,258 patent/US9535767B2/en active Active
-
2010
- 2010-03-19 CA CA3003789A patent/CA3003789C/en active Active
- 2010-03-19 KR KR1020117022331A patent/KR101722276B1/en active IP Right Grant
- 2010-03-19 WO PCT/US2010/028038 patent/WO2010111151A2/en active Application Filing
- 2010-03-19 CA CA2753229A patent/CA2753229C/en active Active
- 2010-03-19 BR BRPI1009274A patent/BRPI1009274A2/en not_active Application Discontinuation
- 2010-03-19 JP JP2012502135A patent/JP5507661B2/en active Active
- 2010-03-19 CN CN2010800137822A patent/CN102365626B/en active Active
- 2010-03-19 RU RU2011139102/08A patent/RU2571366C2/en not_active IP Right Cessation
- 2010-03-19 EP EP10756651.5A patent/EP2411916B1/en active Active
-
2016
- 2016-11-22 US US15/359,561 patent/US10908968B2/en active Active
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20050120160A1 (en) | 2003-08-20 | 2005-06-02 | Jerry Plouffe | System and method for managing virtual servers |
Non-Patent Citations (1)
Title |
---|
See also references of EP2411916A4 |
Also Published As
Publication number | Publication date |
---|---|
KR101722276B1 (en) | 2017-03-31 |
CA2753229C (en) | 2018-10-30 |
US20170075617A1 (en) | 2017-03-16 |
EP2411916A2 (en) | 2012-02-01 |
CA3003789A1 (en) | 2010-09-30 |
CA2753229A1 (en) | 2010-09-30 |
US9535767B2 (en) | 2017-01-03 |
CN102365626A (en) | 2012-02-29 |
EP2411916B1 (en) | 2020-08-19 |
JP5507661B2 (en) | 2014-05-28 |
CN102365626B (en) | 2013-05-29 |
EP2411916A4 (en) | 2013-01-30 |
WO2010111151A3 (en) | 2011-01-13 |
RU2011139102A (en) | 2013-03-27 |
KR20120008026A (en) | 2012-01-25 |
US20100251234A1 (en) | 2010-09-30 |
RU2571366C2 (en) | 2015-12-20 |
US10908968B2 (en) | 2021-02-02 |
CA3003789C (en) | 2020-09-15 |
BRPI1009274A2 (en) | 2016-03-08 |
JP2012521611A (en) | 2012-09-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10908968B2 (en) | Instantiating a virtual machine with a virtual non-uniform memory architecture and determining a highest detected NUMA ratio in a datacenter | |
US10705879B2 (en) | Adjusting guest memory allocation in virtual non-uniform memory architecture (NUMA) nodes of a virtual machine | |
US8701115B2 (en) | Hypervisor scheduler | |
US10691363B2 (en) | Virtual machine trigger | |
US8898664B2 (en) | Exposure of virtual cache topology to a guest operating system | |
US9740517B2 (en) | Dynamic virtual machine memory management | |
CN110447019B (en) | Memory allocation manager and method for managing memory allocation performed thereby |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 201080013782.2 Country of ref document: CN |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 10756651 Country of ref document: EP Kind code of ref document: A2 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2753229 Country of ref document: CA |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2010756651 Country of ref document: EP |
|
ENP | Entry into the national phase |
Ref document number: 20117022331 Country of ref document: KR Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2011139102 Country of ref document: RU |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2012502135 Country of ref document: JP Ref document number: 6935/CHENP/2011 Country of ref document: IN |
|
REG | Reference to national code |
Ref country code: BR Ref legal event code: B01A Ref document number: PI1009274 Country of ref document: BR |
|
ENP | Entry into the national phase |
Ref document number: PI1009274 Country of ref document: BR Kind code of ref document: A2 Effective date: 20110831 |