Typically, in a cloud based computing environment, at least one server assigns workloads (e.g., processes, applications, or other tasks) to one or more computing devices (“managed nodes”) in communication with the server through a network. Some of the managed nodes may be highly occupied with executing workloads that have already been assigned by the server, while others may be only partially occupied or completely unoccupied. By assigning a workload to a managed node that is already heavily loaded with other workloads, the server may cause the managed node to be unable to complete the execution of the assigned workloads in a timely and predictable manner As a result, a customer receiving services from the cloud computing environment may become dissatisfied with the service. On the other hand, performing calculations to assess the capacity of every managed node in the network to accept a workload may be computationally intensive, especially when the cloud based system includes tens of thousands of managed nodes.
The concepts described herein are illustrated by way of example and not by way of limitation in the accompanying figures. For simplicity and clarity of illustration, elements illustrated in the figures are not necessarily drawn to scale. Where considered appropriate, reference labels have been repeated among the figures to indicate corresponding or analogous elements.
While the concepts of the present disclosure are susceptible to various modifications and alternative forms, specific embodiments thereof have been shown by way of example in the drawings and will be described herein in detail. It should be understood, however, that there is no intent to limit the concepts of the present disclosure to the particular forms disclosed, but on the contrary, the intention is to cover all modifications, equivalents, and alternatives consistent with the present disclosure and the appended claims.
References in the specification to “one embodiment,” “an embodiment,” “an illustrative embodiment,” etc., indicate that the embodiment described may include a particular feature, structure, or characteristic, but every embodiment may or may not necessarily include that particular feature, structure, or characteristic. Moreover, such phrases are not necessarily referring to the same embodiment. Further, when a particular feature, structure, or characteristic is described in connection with an embodiment, it is submitted that it is within the knowledge of one skilled in the art to effect such feature, structure, or characteristic in connection with other embodiments whether or not explicitly described. Additionally, it should be appreciated that items included in a list in the form of “at least one A, B, and C” can mean (A); (B); (C); (A and B); (A and C); (B and C); or (A, B, and C). Similarly, items listed in the form of “at least one of A, B, or C” can mean (A); (B); (C); (A and B); (A and C); (B and C); or (A, B, and C).
The disclosed embodiments may be implemented, in some cases, in hardware, firmware, software, or any combination thereof. The disclosed embodiments may also be implemented as instructions carried by or stored on a transitory or non-transitory machine-readable (e.g., computer-readable) storage medium, which may be read and executed by one or more processors. A machine-readable storage medium may be embodied as any storage device, mechanism, or other physical structure for storing or transmitting information in a form readable by a machine (e.g., a volatile or non-volatile memory, a media disc, or other media device).
In the drawings, some structural or method features may be shown in specific arrangements and/or orderings. However, it should be appreciated that such specific arrangements and/or orderings may not be required. Rather, in some embodiments, such features may be arranged in a different manner and/or order than shown in the illustrative figures. Additionally, the inclusion of a structural or method feature in a particular figure is not meant to imply that such feature is required in all embodiments and, in some embodiments, may not be included or may be combined with other features.
The illustrative data center 100 differs from typical data centers in many ways. For example, in the illustrative embodiment, the circuit boards (“sleds”) on which components such as CPUs, memory, and other components are placed are designed for increased thermal performance In particular, in the illustrative embodiment, the sleds are shallower than typical boards. In other words, the sleds are shorter from the front to the back, where cooling fans are located. This decreases the length of the path that air must to travel across the components on the board. Further, the components on the sled are spaced further apart than in typical circuit boards, and the components are arranged to reduce or eliminate shadowing (i.e., one component in the air flow path of another component). In the illustrative embodiment, processing components such as the processors are located on a top side of a sled while near memory, such as DIMMs, are located on a bottom side of the sled. As a result of the enhanced airflow provided by this design, the components may operate at higher frequencies and power levels than in typical systems, thereby increasing performance Furthermore, the sleds are configured to blindly mate with power and data communication cables in each rack 102A, 102B, 102C, 102D, enhancing their ability to be quickly removed, upgraded, reinstalled, and/or replaced. Similarly, individual components located on the sleds, such as processors, accelerators, memory, and data storage drives, are configured to be easily upgraded due to their increased spacing from each other. In the illustrative embodiment, the components additionally include hardware attestation features to prove their authenticity.
Furthermore, in the illustrative embodiment, the data center 100 utilizes a single network architecture (“fabric”) that supports multiple other network architectures including Ethernet and Omni-Path. The sleds, in the illustrative embodiment, are coupled to switches via optical fibers, which provide higher bandwidth and lower latency than typical twister pair cabling (e.g., Category 5, Category 5e, Category 6, etc.). Due to the high bandwidth, low latency interconnections and network architecture, the data center 100 may, in use, pool resources, such as memory, accelerators (e.g., graphics accelerators, FPGAs, ASICs, etc.), and data storage drives that are physically disaggregated, and provide them to compute resources (e.g., processors) on an as needed basis, enabling the compute resources to access the pooled resources as if they were local. The illustrative data center 100 additionally receives usage information for the various resources, predicts resource usage for different types of workloads based on past resource usage, and dynamically reallocates the resources based on this information.
The racks 102A, 102B, 102C, 102D of the data center 100 may include physical design features that facilitate the automation of a variety of types of maintenance tasks. For example, data center 100 may be implemented using racks that are designed to be robotically-accessed, and to accept and house robotically-manipulatable resource sleds. Furthermore, in the illustrative embodiment, the racks 102A, 102B, 102C, 102D include integrated power sources that receive a greater voltage than is typical for power sources. The increased voltage enables the power sources to provide additional power to the components on each sled, enabling the components to operate at higher than typical frequencies.
In various embodiments, dual-mode optical switches may be capable of receiving both Ethernet protocol communications carrying Internet Protocol (IP packets) and communications according to a second, high-performance computing (HPC) link-layer protocol (e.g., Intel's Omni-Path Architecture's, Infiniband) via optical signaling media of an optical fabric. As reflected in
MPCMs 916-1 to 916-7 may be configured to provide inserted sleds with access to power sourced by respective power modules 920-1 to 920-7, each of which may draw power from an external power source 921. In various embodiments, external power source 921 may deliver alternating current (AC) power to rack 902, and power modules 920-1 to 920-7 may be configured to convert such AC power to direct current (DC) power to be sourced to inserted sleds. In some embodiments, for example, power modules 920-1 to 920-7 may be configured to convert 277-volt AC power into 12-volt DC power for provision to inserted sleds via respective MPCMs 916-1 to 916-7. The embodiments are not limited to this example.
MPCMs 916-1 to 916-7 may also be arranged to provide inserted sleds with optical signaling connectivity to a dual-mode optical switching infrastructure 914, which may be the same as—or similar to—dual-mode optical switching infrastructure 514 of
Sled 1004 may also include dual-mode optical network interface circuitry 1026. Dual-mode optical network interface circuitry 1026 may generally comprise circuitry that is capable of communicating over optical signaling media according to each of multiple link-layer protocols supported by dual-mode optical switching infrastructure 914 of
Coupling MPCM 1016 with a counterpart MPCM of a sled space in a given rack may cause optical connector 1016A to couple with an optical connector comprised in the counterpart MPCM. This may generally establish optical connectivity between optical cabling of the sled and dual-mode optical network interface circuitry 1026, via each of a set of optical channels 1025. Dual-mode optical network interface circuitry 1026 may communicate with the physical resources 1005 of sled 1004 via electrical signaling media 1028. In addition to the dimensions of the sleds and arrangement of components on the sleds to provide improved cooling and enable operation at a relatively higher thermal envelope (e.g., 250 W), as described above with reference to
As shown in
In another example, in various embodiments, one or more pooled storage sleds 1132 may be included among the physical infrastructure 1100A of data center 1100, each of which may comprise a pool of storage resources that is available globally accessible to other sleds via optical fabric 1112 and dual-mode optical switching infrastructure 1114. In some embodiments, such pooled storage sleds 1132 may comprise pools of solid-state storage devices such as solid-state drives (SSDs). In various embodiments, one or more high-performance processing sleds 1134 may be included among the physical infrastructure 1100A of data center 1100. In some embodiments, high-performance processing sleds 1134 may comprise pools of high-performance processors, as well as cooling features that enhance air cooling to yield a higher thermal envelope of up to 250 W or more. In various embodiments, any given high-performance processing sled 1134 may feature an expansion connector 1117 that can accept a far memory expansion sled, such that the far memory that is locally available to that high-performance processing sled 1134 is disaggregated from the processors and near memory comprised on that sled. In some embodiments, such a high-performance processing sled 1134 may be configured with far memory using an expansion sled that comprises low-latency SSD storage. The optical infrastructure allows for compute resources on one sled to utilize remote accelerator/FPGA, memory, and/or SSD resources that are disaggregated on a sled located on the same rack or any other rack in the data center. The remote resources can be located one switch jump away or two-switch jumps away in the spine-leaf network architecture described above with reference to
In various embodiments, one or more layers of abstraction may be applied to the physical resources of physical infrastructure 1100A in order to define a virtual infrastructure, such as a software-defined infrastructure 1100B. In some embodiments, virtual computing resources 1136 of software-defined infrastructure 1100B may be allocated to support the provision of cloud services 1140. In various embodiments, particular sets of virtual computing resources 1136 may be grouped for provision to cloud services 1140 in the form of SDI services 1138. Examples of cloud services 1140 may include—without limitation—software as a service (SaaS) services 1142, platform as a service (PaaS) services 1144, and infrastructure as a service (IaaS) services 1146.
In some embodiments, management of software-defined infrastructure 1100B may be conducted using a virtual infrastructure management framework 1150B. In various embodiments, virtual infrastructure management framework 1150B may be designed to implement workload fingerprinting techniques and/or machine-learning techniques in conjunction with managing allocation of virtual computing resources 1136 and/or SDI services 1138 to cloud services 1140. In some embodiments, virtual infrastructure management framework 1150B may use/consult telemetry data in conjunction with performing such resource allocation. In various embodiments, an application/service management framework 1150C may be implemented in order to provide QoS management capabilities for cloud services 1140. The embodiments are not limited in this context.
As shown in
Each managed node 1260, in the illustrative embodiment, continually performs a self-evaluation as the managed node 1260 executes one or more workloads to determine whether the managed node 1260 is able to take on an additional workload. In doing so, each managed node 1260 generates telemetry data indicative of performance and conditions (e.g., resource utilization, one or more temperatures, fan speeds, etc.) as the managed node 1260 executes one or more workloads and compares the telemetry data to predefined thresholds. If the values in the telemetry data satisfy the thresholds (e.g., a present processor utilization is less than a predefined threshold processor utilization), the managed node 1260 determines that it is available for an additional workload. Otherwise, the managed node 1260 determines that it is unavailable for an additional workload. In the illustrative embodiment, the predefined thresholds may vary, depending on whether the managed node 1260 has been assigned a workload that is to be executed with deterministic (i.e., predictable) performance (e.g., high priority) rather than a normal priority. As such, when executing a workload that has been designated to be executed deterministically, the processor utilization threshold may be a lower value (e.g., 70%) than the processor utilization threshold (e.g., 80%) if the managed node 1260 is executing workloads that do not have high priority. Furthermore, the managed nodes 1260 may communicate with each other to collect availability data from other managed nodes 1260, such as with a bee foraging algorithm, to identify the managed nodes 1260 available to receive additional workloads, rather than each managed node 1260 independently reporting its availability data directly to the orchestrator server 1240.
Referring now to
The CPU 1302 may be embodied as any type of processor capable of performing the functions described herein. The CPU 1302 may be embodied as a single or multi-core processor(s), a microcontroller, or other processor or processing/controlling circuit. In some embodiments, the CPU 1302 may be embodied as, include, or be coupled to a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), reconfigurable hardware or hardware circuitry, or other specialized hardware to facilitate performance of the functions described herein. As discussed above, the managed node 1260 may include resources distributed across multiple sleds and in such embodiments, the CPU 1302 may include portions thereof located on the same sled or different sled. Similarly, the main memory 1304 may be embodied as any type of volatile (e.g., dynamic random access memory (DRAM), etc.) or non-volatile memory or data storage capable of performing the functions described herein. In some embodiments, all or a portion of the main memory 1304 may be integrated into the CPU 1302. In operation, the main memory 1304 may store various software and data used during operation such as availability data, telemetry data, policy data, workload labels, workload classifications, workload adjustment data, operating systems, applications, programs, libraries, and drivers. As discussed above, the managed node 1260 may include resources distributed across multiple sleds and in such embodiments, the main memory 1304 may include portions thereof located on the same sled or different sled.
The I/O subsystem 1306 may be embodied as circuitry and/or components to facilitate input/output operations with the CPU 1302, the main memory 1304, and other components of the orchestrator server 1240. For example, the I/O subsystem 1306 may be embodied as, or otherwise include, memory controller hubs, input/output control hubs, integrated sensor hubs, firmware devices, communication links (e.g., point-to-point links, bus links, wires, cables, light guides, printed circuit board traces, etc.), and/or other components and subsystems to facilitate the input/output operations. In some embodiments, the I/O subsystem 1306 may form a portion of a system-on-a-chip (SoC) and be incorporated, along with one or more of the CPU 1302, the main memory 1304, and other components of the orchestrator server 1240, on a single integrated circuit chip.
The communication circuitry 1308 may be embodied as any communication circuit, device, or collection thereof, capable of enabling communications over the network 1230 between the orchestrator server 1240 and another compute device (e.g., the client device 1220 and/or the managed nodes 1260). The communication circuitry 1308 may be configured to use any one or more communication technology (e.g., wired or wireless communications) and associated protocols (e.g., Ethernet, Bluetooth®, Wi-Fi®, WiMAX, etc.) to effect such communication.
The illustrative communication circuitry 1308 includes a network interface controller (NIC) 1310, which may also be referred to as a host fabric interface (HFI). The NIC 1310 may be embodied as one or more add-in-boards, daughtercards, network interface cards, controller chips, chipsets, or other devices that may be used by the orchestrator server 1240 to connect with another compute device (e.g., a managed node 1260 or the client device 1220). In some embodiments, the NIC 1310 may be embodied as part of a system-on-a-chip (SoC) that includes one or more processors, or included on a multichip package that also contains one or more processors. In some embodiments, the NIC 1310 may include a local processor (not shown) and/or a local memory (not shown) that are both local to the NIC 1310. In such embodiments, the local processor of the NIC 1310 may be capable of performing one or more of the functions of the CPU 1302 described herein. Additionally or alternatively, in such embodiments, the local memory of the NIC 1310 may be integrated into one or more components of the orchestrator server 1240 at the board level, socket level, chip level, and/or other levels. As discussed above, the managed node 1260 may include resources distributed across multiple sleds and in such embodiments, the communication circuitry 1308 may include portions thereof located on the same sled or different sled.
The one or more illustrative data storage devices 1312, may be embodied as any type of devices configured for short-term or long-term storage of data such as, for example, memory devices and circuits, memory cards, hard disk drives, solid-state drives, or other data storage devices. Each data storage device 1312 may include a system partition that stores data and firmware code for the data storage device 1312. Each data storage device 1312 may also include an operating system partition that stores data files and executables for an operating system.
Additionally, the orchestrator server 1240 may include a display 1314. The display 1314 may be embodied as, or otherwise use, any suitable display technology including, for example, a liquid crystal display (LCD), a light emitting diode (LED) display, a cathode ray tube (CRT) display, a plasma display, and/or other display usable in a compute device. The display 1314 may include a touchscreen sensor that uses any suitable touchscreen input technology to detect the user's tactile selection of information displayed on the display including, but not limited to, resistive touchscreen sensors, capacitive touchscreen sensors, surface acoustic wave (SAW) touchscreen sensors, infrared touchscreen sensors, optical imaging touchscreen sensors, acoustic touchscreen sensors, and/or other type of touchscreen sensors.
Additionally or alternatively, the orchestrator server 1240 may include one or more peripheral devices 1316. Such peripheral devices 1316 may include any type of peripheral device commonly found in a compute device such as speakers, a mouse, a keyboard, and/or other input/output devices, interface devices, and/or other peripheral devices.
The client device 1220 and the managed nodes 1260 may have components similar to those described in
As described above, the client device 1220, the orchestrator server 1240 and the managed nodes 1260 are illustratively in communication via the network 1230, which may be embodied as any type of wired or wireless communication network, including global networks (e.g., the Internet), local area networks (LANs) or wide area networks (WANs), cellular networks (e.g., Global System for Mobile Communications (GSM), 3G, Long Term Evolution (LTE), Worldwide Interoperability for Microwave Access (WiMAX), etc.), digital subscriber line (DSL) networks, cable networks (e.g., coaxial networks, fiber networks, etc.), or any combination thereof.
Referring now to
Additionally, the illustrative environment 1400 includes policy data 1404 indicative of user-defined preferences as to the heat production, power consumption, and life expectancy of the components of the managed nodes 1260. Further, the illustrative environment 1400 includes workload labels 1406 which may be embodied as any identifiers (e.g., process numbers, executable file names, alphanumeric tags, etc.) that uniquely identify each workload executed by the managed nodes 1260. In addition, the illustrative environment 1400 includes workload classifications 1408 which may be embodied as any data indicative of the resource utilization tendencies of each workload (e.g., processor intensive, network bandwidth intensive, etc.). Further, the illustrative environment 1400 includes workload adjustment data 1410 which may be embodied as any data indicative of reassignments (e.g., live migrations) of one or more workloads from one managed node 1260 to another managed node 1260 and/or adjustments to settings for components within each managed node 1260, such as processor capacity (e.g., a number of cores to be used, a clock speed, a percentage of available processor cycles, etc.) available to one or more workloads, memory resource capacity (e.g., amount of memory to be used and/or frequency of memory accesses to volatile memory and/or non-volatile memory) available to one or more workloads, and/or communication circuitry capacity available to one or more workloads. The illustrative embodiment additionally includes availability data 1412, which may be embodied as any data indicative of a determination made by each of the managed nodes 1260 as to whether the managed node 1260 is able to receive and execute another workload. In the illustrative embodiment, the orchestrator server 1240 continually receives updated availability data 1412 such that a particular managed node 1260 that initially reported an unavailability to take on an additional workload may later report that it is able to execute an additional workload. As described herein, the managed nodes 1260 that reported an availability to perform additional workloads form a “short list” of managed nodes 1260 to be analyzed in more detail by the orchestrator server 1240.
In the illustrative environment 1400, the network communicator 1420, which may be embodied as hardware, firmware, software, virtualized hardware, emulated architecture, and/or a combination thereof as discussed above, is configured to facilitate inbound and outbound network communications (e.g., network traffic, network packets, network flows, etc.) to and from the orchestrator server 1240, respectively. To do so, the network communicator 1420 is configured to receive and process data packets from one system or computing device (e.g., the client device 1220) and to prepare and send data packets to another computing device or system (e.g., the managed nodes 1260). Accordingly, in some embodiments, at least a portion of the functionality of the network communicator 1420 may be performed by the communication circuitry 1308, and, in the illustrative embodiment, by the NIC 1310.
The telemetry monitor 1430, which may be embodied as hardware, firmware, software, virtualized hardware, emulated architecture, and/or a combination thereof as discussed above, is configured to collect status data (e.g., telemetry data 1402 and managed node availability data 1412) from the managed nodes 1260 as the managed nodes 1260 execute the workloads assigned to them. The telemetry monitor 1430 may actively poll each of the managed nodes 1260 for updated status data on an ongoing basis or may passively receive the status data from the managed nodes 1260, such as by listening on a particular network port for updated status data. The telemetry monitor 1430 may further parse and categorize the status data, such as by separating the status data into an individual file or data set for each managed node 1260. In the illustrative embodiment, the telemetry monitor 1430 includes a node availability data collector 1432 to receive and parse the availability data 1412 for each of the managed nodes 1260. The node availability data collector 1432, in the illustrative embodiment, may receive availability data 1412 from one or more managed nodes 1260 on behalf of multiple other managed nodes 1260, rather than receiving the availability data directly from each managed node 1260. In such embodiments, the node availability data collector 1432 may parse an aggregated set of availability data 1412 received from one of the managed nodes 1260 to identify which portions of the availability data 1412 pertain to which managed nodes 1260. The node availability data collector 1432 may also overwrite earlier availability data for a particular managed node 1260 with updated availability data 1412, compare a present time to a time stamp associated with existing availability data 1412 from a managed node 1260 to determine whether the availability data 1412 is potentially outdated (i.e., older than a predefined time period), and, in response to a determination that the availability data 1412 is potentially outdated, prompt the corresponding managed nodes 1260 for updated availability data 1412.
The policy manager 1440, which may be embodied as hardware, firmware, software, virtualized hardware, emulated architecture, and/or a combination thereof as discussed above, is configured to receive and store the policy data 1404, which, as described above, is indicative of user-defined preferences pertaining to operating parameters of the components of the managed nodes 1260 that may affect, among other items, heat production, power consumption, and/or life expectancy (i.e., wear) of the managed nodes 1260. The policy manager 1440 is further configured to provide the policy data 1404 to the resource manager 1450 to assist in determining adjustments to the assignment of workloads among the managed nodes 1260 and for adjusting settings within one or more of the managed nodes (e.g., processor capacity available to one or more workloads, memory resource capacity available to one or more workloads, and/or communication circuitry capacity available to one or more workloads) to optimize resource utilization, subject to the policies defined in the policy data 1404.
The resource manager 1450, which may be embodied as hardware, firmware, software, virtualized hardware, emulated architecture, and/or a combination thereof, is configured to generate data analytics from the telemetry data 1402, identify the workloads, classify the workloads, identify trends in the resource utilization of the workloads, predict future resource utilizations of the workloads, and adjust the assignments of the workloads to the managed nodes 1260 and the settings of the managed nodes 1260 to increase the resource utilization (e.g., to reduce the amount of idle resources) while staying in compliance with the policy data 1404. For efficiency, in the illustrative embodiment, the resource manager 1450 limits the above analysis to the managed nodes 1260 that reported an availability to receive an additional workload, thereby significantly reducing the computational burden on the orchestrator server 1240 in assigning and balancing workloads across the managed nodes 1260. In the illustrative embodiment, the resource manager 1450 includes an analysis limiter 1452, a workload labeler 1454, a workload classifier 1456, a workload behavior predictor 1458, a workload placer 1460, and a node settings adjuster 1462. The analysis limiter 1452, in the illustrative embodiment, is configured to analyze the availability data 1412 and generate, as a function of the availability data, a “short list” (i.e., a reduced set) of the managed nodes 1260 for analysis by the workload labeler 1454, the workload classifier 1456, the workload behavior predictor 1458, the workload placer 1460, and the node settings adjuster 1462. In the illustrative embodiment, the analysis limiter 1452 adds to the reduced set, identifiers of the managed nodes 1260 that indicated, in the availability data 1412, that they are available to receive an additional workload and excludes the managed nodes 1260 that indicated an unavailability to receive an additional workload.
The workload labeler 1454, in the illustrative embodiment, is configured to assign a workload label 1406 to each workload presently performed or scheduled to be performed by one or more of the managed nodes 1260 in the reduced set. The workload labeler 1454 may generate the workload label 1406 as a function of an executable name of the workload, a hash of all or a portion of the code of the workload, or based on any other method to uniquely identify each workload. The workload classifier 1456, in the illustrative embodiment, is configured to categorize each labeled workload based on the resource utilization usage of each workload. For example, the workload classifier 1456 may categorize one set of labeled workloads as being consistently processor intensive, another set of labeled workloads as being consistently memory intensive, and another set of workloads as having phases of different resource utilization (high memory use and low processor use, followed by high processor use and low memory use, etc.).
The workload behavior predictor 1458, in the illustrative embodiment, is configured to analyze the telemetry data 1402 and the workload classifications 1408 to predict future resource utilization needs of the various workloads based on their previous usage. In doing so, the workload behavior predictor 1458 may determine a present phase of a given workload and determine an amount of remaining time until the workload transitions to another phase having different resource utilization characteristics. The workload placer 1460, in the illustrative embodiment, is configured to initially assign workloads to the various managed nodes 1260 in the reduced set generated by the analysis limiter 1452, and determine, based on the telemetry data 1402, the workload classifications 1408, and the policy data 1404, whether the resources of the managed nodes 1260 could be more efficiently used (e.g., to reduce the amount of idle resources and to reduce the load on over-used resources) by reassigning the workloads among the managed nodes 1260, without violating the policies in the policy data (e.g., without generating more than a threshold amount of heat, without consuming more than a threshold amount of power, etc.). Similarly, the node settings adjuster 1462, in the illustrative embodiment, is configured to determine one or more adjustments to the settings within the reduced set of managed nodes 1260 to provide or restrict the resources available to the workloads in accordance with the goal of optimizing resource usage and maintaining conformance with the policies in the policy data 1404. The settings may be associated with the operating system and/or the firmware or drivers of the components of the managed nodes 1260.
It should be appreciated that each of the analysis limiter 1452, workload labeler 1454, the workload classifier 1456, the workload behavior predictor 1458, the workload placer 1460, and the node settings adjuster 1462 may be separately embodied as hardware, firmware, software, virtualized hardware, emulated architecture, and/or a combination thereof. For example, the analysis limiter 1452 may be embodied as a hardware component, while the workload labeler 1454, the workload classifier 1456, the workload behavior predictor 1458, the workload placer 1460, and the node settings adjuster 1462 are embodied as a virtualized hardware component or as some other combination of hardware, firmware, software, virtualized hardware, emulated architecture, and/or a combination thereof. Each of the components of the environment 1400 may be embodied as hardware, firmware, software, or a combination thereof.
Referring now to
In the illustrative environment 1500, the network communicator 1520, which may be embodied as hardware, firmware, software, virtualized hardware, emulated architecture, and/or a combination thereof as discussed above, is configured to facilitate inbound and outbound network communications (e.g., network traffic, network packets, network flows, etc.) to and from the managed node 1260, respectively. To do so, the network communicator 1520 is configured to receive and process data packets from one system or computing device (e.g., the client device 1220, the orchestrator server 1240, and/or another managed node 1260) and to prepare and send data packets to another computing device or system (e.g., the client device 1220, the orchestrator server 1240, and/or one another managed node 1260). Accordingly, in some embodiments, at least a portion of the functionality of the network communicator 1520 may be performed by the communication circuitry 1308, and, in the illustrative embodiment, by the NIC 1310.
The workload executor 1530, which may be embodied as hardware, firmware, software, virtualized hardware, emulated architecture, and/or a combination thereof as discussed above, is configured to execute workloads assigned to the managed node 1260. The telemetry data generator 1540, which may be embodied as hardware, firmware, software, virtualized hardware, emulated architecture, and/or a combination thereof as discussed above, is configured to monitor the performance and conditions within the managed node 1260 as the one or more workloads are executed and generate the telemetry data 1506.
The availability data manager 1550, which may be embodied as hardware, firmware, software, virtualized hardware, emulated architecture, and/or a combination thereof as discussed above, is configured to generate the availability data 1508 and report the availability data 1508 either directly to the orchestrator server 1240 or to another managed node 1260. The availability data manager 1550 may additionally aggregate the availability data 1508 from one or more other managed nodes 1260, such as managed nodes 1260 having a predefined relationship to the managed node 1260 (e.g., within a predefined logical proximity of the managed node 1260, such as on the same network switch), identified in a predefined set of managed nodes 1260 from which to collect the availability data 1508, or identified as managed nodes 1260 to collect the availability data 1508 from, pursuant to a swarm intelligence algorithm, such as a bee foraging algorithm. To do so, in the illustrative embodiment, the availability data manager 1550 includes an availability data determiner 1552, an availability data reporter 1554, and an availability data aggregator 1556.
The availability data determiner 1552, in the illustrative embodiment, is configured to compare resource utilization values (e.g., processor utilization, memory utilization, network bandwidth utilization, etc.) in the telemetry data 1506 to a set of predefined threshold values such as a processor utilization threshold, a memory usage threshold, and/or a network bandwidth threshold to determine an availability of the managed node 1260 to receive and execute an additional workload. Accordingly, if one or more of the existing utilizations of one or more of the resources in the managed node 1260 is in excess of a corresponding predefined threshold, the availability data determiner 1552 may store, in the availability data, an indication that the managed node 1260 is presently unavailable to execute an additional workload. Otherwise, the availability data determiner 1552 may store an indication that the managed node 1260 is presently available to execute an additional workload. Furthermore, in the illustrative embodiment, the availability data determiner 1552 may select one of multiple sets of predefined threshold values as a function of the priorities assigned to the existing workloads. In the illustrative embodiment, if one or more of the existing workloads has a high priority, meaning the workload is to be executed at a predictable speed, the availability data determiner 1552 may select a set of corresponding predefined thresholds with lower resource utilization values than if none of the workloads have been designated as high priority. Doing so may protect high priority workloads from possible interruption from additional workloads, while enabling managed nodes 1260 without high priority workloads to take on additional work.
The availability data reporter 1554, in the illustrative embodiment, is configured to report the availability data 1508 to the orchestrator server 1240, either directly or through another managed node 1260. The availability data reporter 1554 may report the availability data 1508 on a repeating, periodic basis without prompting from another compute device, or may report the availability data 1508 in response to a query from the orchestrator server 1240 or another managed node 1260. The availability data aggregator 1556, in the illustrative embodiment, is configured to aggregate availability data 1508 from at least one other managed node 1260. In doing so, the availability data aggregator may receive the availability data 1508 from one or more managed nodes 1260 that have a predefined relationship to the present managed node 1260, that are listed in a predefined set of managed nodes 1260 from which to receive availability data 1508, or that are otherwise identified to the managed node 1260, such as pursuant to a swarm intelligence algorithm. In a swarm intelligence algorithm, the availability data aggregator 1556 may determine that one or more managed nodes 1260 are within an “area” (e.g., a set of managed nodes 1260) that has historically been available to take on additional workloads. As such, the managed nodes 1260 within such areas are more frequently checked for their availability to execute additional workloads. In some embodiments, the availability data aggregator 1556 may provide identifiers of managed nodes 1260 in such an area to other managed nodes 1260 that are responsible for aggregating and reporting back availability data to the orchestrator server 1240. In response, those managed nodes 1260 may frequently check the availability of managed nodes 1260 in that area and/or other nearby managed nodes 1260 (e.g., within the same rack, connected to the same switch, or otherwise within a predefined range from a physical or network topology perspective). As such, the managed nodes 1260 may exhibit a swarm intelligence when identifying sets of managed nodes 1260 available to perform additional workloads.
It should be appreciated that each of the availability data determiner 1552, the availability data reporter 1554, and the availability data aggregator 1556 may be separately embodied as hardware, firmware, software, virtualized hardware, emulated architecture, and/or a combination thereof. For example, the availability data determiner 1552 may be embodied as a hardware component, while the availability data reporter 1554 and the availability data aggregator 1556 are embodied as a virtualized hardware component or as some other combination of hardware, firmware, software, virtualized hardware, emulated architecture, and/or a combination thereof. Each of the components of the environment 1500 may be embodied as hardware, firmware, software, or a combination thereof.
Referring now to
After receiving the policy data 1404, in the illustrative embodiment, the method 1600 advances to block 1606 in which the orchestrator server 1240 assigns initial workloads to the managed nodes 1260. In the illustrative embodiment, the orchestrator server 1240 has not received telemetry data 1402 that would inform a decision as to where the workloads are to be assigned among the managed nodes 1260. As such, the orchestrator server 1240 may assign the workloads to the managed nodes 1260 based on any suitable method, such as assigning each workload to the first available managed node that is idle (i.e., is not presently executing a workload), randomly assigning the workloads, or by any other method. In the illustrative embodiment, as indicated in block 1606, in assigning the initial workloads to the managed nodes 1260, the orchestrator server 1240 may assign a priority to each of the workloads, such as by storing an indicator of the priority in data describing each workload (e.g., the workload data 1504). In doing so, the orchestrator server 1240 may assign a normal priority to one or more of the workloads, as indicated in block 1610. In the illustrative embodiment, a normal priority is a priority in which the workload is not required to produce output at specific instances in time. Alternatively, as indicated in block 1612, the orchestrator server 1240 may assign a deterministic execution priority (i.e., a high priority) to one or more of the workloads, indicating that the workload is to be executed in a predictable manner and produce outputs at specific times. The priorities may be determined based on input from the client device 1220, such as a selection of the desired responsiveness and speed of the services to be provided by the system 1210. In the illustrative embodiment, the orchestrator server 1240 may generate initial availability data based on the assignment of the workloads among the managed nodes 1260, as indicated in block 1614. In doing so, the orchestrator server 1240 may estimate an expected amount of resources that will be consumed by each workload, based on the priorities associated with the workloads and/or based on previously generated profiles (e.g., workload classifications 1408) if such data is presently available to the orchestrator server 1240.
After assigning the initial workloads to the managed nodes 1260, the method 1600 advances to block 1616 in which the orchestrator server 1240 receives status data from the managed nodes 1260 as the workloads are performed (i.e., executed). In receiving the status data, the orchestrator server 1240 receives the availability data 1412 from one or more of the managed nodes 1260 indicating the availability of each managed node 1260 to receive and perform an additional workload, as represented in block 1618. Further, in receiving the availability data 1412, the orchestrator server 1240, in the illustrative embodiment, determines a reduced set of available nodes from the availability data 1412. In the illustrative embodiment, the reduced set of available nodes is the subset of the managed nodes 1260 that reported that they are available to receive and execute an additional workload. Additionally, in receiving the status data, the orchestrator server 1240 receives the telemetry data 1402 from the managed nodes 1260 as the workloads are performed (i.e., executed), as indicated in block 1622. In doing so, the orchestrator server 1240 may receive temperature data indicative of a temperature within each managed node 1260, power consumption data indicative of an amount of power consumed by each managed node 1260, processor utilization data indicative of an amount of processor usage consumed by each workload performed by each managed node 1260, memory utilization data for each managed node 1260 (cache utilization data, other volatile memory utilization, and/or non-volatile memory utilization), network utilization data indicative of an amount of network bandwidth used by each workload performed by each managed node 1260, and/or data indicative of other conditions within each managed node 1260. After receiving the status data, the orchestrator server 1240 generates data analytics, as described below.
Referring now to
In block 1638, the orchestrator server 1240 determines, as a function of the data analytics, adjustments to the workload assignments as the workloads are performed, to improve resource utilization. In block 1640, the orchestrator server 1240 may add or change workload assignments among the managed nodes 1260. In doing so, the orchestrator server 1240 may identify one or more available managed nodes 1260 executing workloads with relatively low resource utilization and assign additional workloads to those managed nodes 1260. As stated above, the orchestrator may also reassign workloads among the managed nodes 1260. For example, the orchestrator server 1240 may identify, based on the data analytics, workloads having complementary resource utilizations (e.g., a workload with a high processor utilization and low memory utilization and another workload with low processor utilization and high memory utilization), and assign those two workloads to the same managed node 1260 to improve the resource utilization. In the illustrative embodiment, the orchestrator server 1240 limits the additions and changes to the workload assignments to only the reduced set of available managed nodes 1260.
The orchestrator server 1240 may additionally determine node-specific adjustments, as indicated in block 1644. The node-specific adjustments may be embodied as changes to settings within one or more of the managed nodes 1260, such as in the operating system, the drivers, and/or the firmware of components (e.g., the CPU 1302, the memory 1304, the communication circuitry 1308, the one or more data storage devices 1312, etc.) to improve resource utilization. As such, in the illustrative embodiment, in determining the node-specific adjustments, the orchestrator server 1240 may determine processor throttle adjustments, such as clock speed and/or processor affinity for one or more workloads, memory usage adjustments, such as allocations of volatile memory (e.g., the memory 1304) and/or data storage capacity (e.g., capacity of the one or more data storage devices 1312), memory bus speeds, and/or other memory-related settings, network bandwidth adjustments, such as an available bandwidth of the communication circuitry 1308 to be allocated to each workload, and/or one or more fan speed adjustments to increase or decrease the cooling within the managed node 1260. In doing so, in the illustrative embodiment, the orchestrator server 1240 limits the node-specific adjustments to the reduced set of available managed nodes 1260. In block 1564, the orchestrator server 1240 may modify the adjustments to the assignments of the workloads and/or to the node-specific adjustments to comply with the policy data 1404. As an example, the policy data 1404 may indicate that the power consumption is not to exceed a predefined threshold and, in view of the threshold, the orchestrator server 1240 may determine to reduce the speed of the CPU 1302 to satisfy the threshold and reassign a processor-intensive workload away from the managed node 1260 because, at the reduced speed, the CPU 1302 would be unable to complete the processor-intensive workload within a predefined time period (e.g., a time period specified in a Service Level Agreement (SLA) between the user of the client device 1220 and the operator of the system 1210).
Referring now to
Referring now to
After receiving the workload assignments, the managed node 1260 may receive node-specific adjustments from the orchestrator server 1240, such as changes to settings in the operating system, the drivers, and/or the firmware of components (e.g., the CPU 1302, the memory 1304, the communication circuitry 1308, the one or more data storage devices 1312, etc.) to alter the power and/or resource utilization of the managed node 1260. In block 1916, the managed node 1260 executes the assigned workload. In doing so, the managed node 1260 may apply the node-specific adjustments received in block 1914. Subsequently, as indicated in block 1920, the managed node 1260 may receive a request for availability data. In receiving the request for availability data, the managed node 1260 may receive the request from the orchestrator server 1240 as indicated in block 1922. Alternatively, the managed node 1260 may receive the request from another managed node 1260, as indicated in block 1924.
Referring now to
After the managed node 1260 generates the telemetry data 1506, the method 1900 advances to block 1938, in which the managed node 1260 compares the telemetry data 1506 to one or more predefined thresholds to determine an availability of the managed node 1260 to receive and execute an additional workload. In doing so, the managed node 1260 may select a set of predefined thresholds as a function of the indication of the priority of the workload (e.g., an indication of the priority in the workload data 1504). For example, if an assigned workload has been designated as high priority (e.g., to be executed deterministically) the managed node 1260 may select a set of predefined thresholds with lower values that, if exceeded, would cause the managed node 1260 to be deemed unavailable to take on an additional workload. As such, the processor utilization threshold when the managed node 1260 is executing a high priority workload may be a lower value (e.g., 70%) than the processor utilization threshold (e.g., 80%) if the managed node 1260 is presently only executing workloads that do not have high priority. As indicated in block 1942, the managed node 1260 may compare the processor utilization to a predefined processor availability threshold. Additionally or alternatively, the managed node 1260 may compare the memory utilization data to a predefined memory availability threshold, as indicated in block 1944, and/or may compare other components of the telemetry data 1506 to corresponding availability thresholds (e.g., a predefined network bandwidth availability threshold, a predefined power consumption availability threshold, a predefined temperature availability threshold, etc.), as indicated in block 1946.
In block 1948, the managed node 1260 determines whether the thresholds were satisfied. In the illustrative embodiment, if any of the values in the telemetry data 1506 exceeded a corresponding predefined threshold, the managed node 1260 determines that the thresholds were not satisfied. In other embodiments, the managed node 1260 may determine whether the thresholds were satisfied based on another scheme (e.g., whether a majority of the predefined thresholds were exceeded, etc.). Regardless, in response to a determination that the thresholds were not satisfied, the method 1900 advances to block 1950 in which the managed node 1260 stores an indication of non-availability in the availability data 1508. Otherwise, the method 1900 advances to block 1952, in which the managed node 1260 stores an indication that the managed node 1260 is available in the availability data 1508. In either case, the method 1900 proceeds with the collection and reporting of the availability data 1508 to the orchestrator server 1240, as described herein.
Referring now to
In block 1964, the managed node 1260 reports status data. In doing so, as indicated in block 1966, the managed node reports the availability data 1508. In reporting the availability data, the managed node 1260 may report the availability data to the orchestrator server 1240 directly, as indicated in block 1968. Alternatively, the managed node 1260 may report the availability data to another managed node 1260 to be collected (i.e., aggregated) and reported back to the orchestrator server 1240. In block 1974, the managed node 1260 also reports the telemetry data 1506 to the orchestrator server 1240. After the managed node 1260 has reported the status data, the method 1900 loops back to block 1902 in which the managed node 1260 determines whether to continue operations (i.e., to repeat the method 1900).
Illustrative examples of the technologies disclosed herein are provided below. An embodiment of the technologies may include any one or more, and any combination of, the examples described below.
Example 1 includes an orchestrator server to utilize availability data for a set of managed nodes to assign workloads, the orchestrator server comprising one or more processors; one or more memory devices having stored therein a plurality of instructions that, when executed by the one or more processors, cause the orchestrator server to assign workloads to the managed nodes; receive availability data from the managed nodes, wherein the availability data is indicative of a determination by each of the managed nodes as to an availability of the managed node to receive an additional workload; receive telemetry data from the managed nodes, wherein the telemetry data is indicative of resource utilization by each of the managed nodes as the workloads are performed; determine, as a function of the availability data, a reduced set of available managed nodes for analysis; determine, as a function of the telemetry data, adjustments to the workload assignments to increase the resource utilization among the reduced set of managed nodes; and apply the determined adjustments to the reduced set of managed nodes as the workloads are performed.
Example 2 includes the subject matter of Example 1, and wherein to assign the workloads comprises to assign a priority to one or more of the workloads.
Example 3 includes the subject matter of any of Examples 1 and 2, and wherein to assign a priority to one or more of the workloads comprises to assign a deterministic execution priority to one or more of the workloads.
Example 4 includes the subject matter of any of Examples 1-3, and wherein to assign the workloads comprises to generate availability data as a function of the assignment of the workloads.
Example 5 includes the subject matter of any of Examples 1-4, and wherein to determine, as a function of the telemetry data, adjustments to the workload assignments comprises to generate, as a function of the telemetry data, data analytics as the workloads are performed.
Example 6 includes the subject matter of any of Examples 1-5, and wherein to generate data analytics comprises to limit the generation of the data analytics to the reduced set of managed nodes.
Example 7 includes the subject matter of any of Examples 1-6, and wherein to generate data analytics comprises to identify trends in resource utilization of the workloads performed by the managed nodes in the reduced set of managed nodes.
Example 8 includes the subject matter of any of Examples 1-7, and wherein to generate data analytics comprises to generate profiles of the workloads performed by the managed nodes in the reduced set of managed nodes.
Example 9 includes the subject matter of any of Examples 1-8, and wherein to generate data analytics comprises to predict future resource utilization of the workloads performed by the managed nodes in the reduced set of managed nodes.
Example 10 includes the subject matter of any of Examples 1-9, and wherein the plurality of instructions, when executed by the one or more processors, further the cause the orchestrator server to obtain policy data indicative of one or more goals to be achieved in the management of the workloads; and modify the adjustments as a function of the policy data.
Example 11 includes the subject matter of any of Examples 1-10, and wherein to determine the adjustments comprises to determine one or more node-specific adjustments indicative of changes to an availability of one or more resources of a managed node in the reduced set of managed nodes to one or more of the workloads performed by the managed node.
Example 12 includes the subject matter of any of Examples 1-11, and wherein to determine the node-specific adjustments comprises to determine at least one of a processor throttle adjustment, a memory usage adjustment, a network bandwidth adjustment, or a fan speed adjustment.
Example 13 includes the subject matter of any of Examples 1-12, and wherein to apply the determined adjustments comprises to issue a request to perform a live migration of a workload between the managed nodes.
Example 14 includes the subject matter of any of Examples 1-13, and wherein to apply the determined adjustments comprises to issue a request to one of the managed nodes to apply one or more node-specific adjustments indicative of changes to an availability of one or more resources of the managed node to one or more of the workloads performed by the managed node.
Example 15 includes a method for utilizing availability data for a set of managed nodes to assign workloads, the method comprising assigning, by an orchestrator server, workloads to the managed nodes; receiving, by the orchestrator server, availability data from the managed nodes, wherein the availability data is indicative of a determination by each of the managed nodes as to an availability of the managed node to receive an additional workload; receiving, by the orchestrator server, telemetry data from the managed nodes, wherein the telemetry data is indicative of resource utilization by each of the managed nodes as the workloads are performed; determining, by the orchestrator server and as a function of the availability data, a reduced set of available managed nodes for analysis; determining, by the orchestrator server and as a function of the telemetry data, adjustments to the workload assignments to increase the resource utilization among the reduced set of managed nodes; and applying, by the orchestrator server, the determined adjustments to the reduced set of managed nodes as the workloads are performed.
Example 16 includes the subject matter of Example 15, and wherein assigning the workloads comprises assigning a priority to one or more of the workloads.
Example 17 includes the subject matter of any of Examples 15 and 16, and wherein assigning a priority to one or more of the workloads comprises assigning a deterministic execution priority to one or more of the workloads.
Example 18 includes the subject matter of any of Examples 15-17, and wherein assigning the workloads comprises generating availability data as a function of the assignment of the workloads.
Example 19 includes the subject matter of any of Examples 15-18, and wherein determining, as a function of the telemetry data, adjustments to the workload assignments comprises generating, as a function of the telemetry data, data analytics as the workloads are performed.
Example 20 includes the subject matter of any of Examples 15-19, and wherein generating data analytics comprises limiting the generation of the data analytics to the reduced set of managed nodes.
Example 21 includes the subject matter of any of Examples 15-20, and wherein generating data analytics comprises identifying trends in resource utilization of the workloads performed by the managed nodes in the reduced set of managed nodes.
Example 22 includes the subject matter of any of Examples 15-21, and wherein generating data analytics comprises generating profiles of the workloads performed by the managed nodes in the reduced set of managed nodes.
Example 23 includes the subject matter of any of Examples 15-22, and wherein generating data analytics comprises predicting future resource utilization of the workloads performed by the managed nodes in the reduced set of managed nodes.
Example 24 includes the subject matter of any of Examples 15-23, and further including obtaining, by the orchestrator server, policy data indicative of one or more goals to be achieved in the management of the workloads; and modifying, by the orchestrator server, the adjustments as a function of the policy data.
Example 25 includes the subject matter of any of Examples 15-24, and wherein determining the adjustments comprises determining one or more node-specific adjustments indicative of changes to an availability of one or more resources of a managed node in the reduced set of managed nodes to one or more of the workloads performed by the managed node.
Example 26 includes the subject matter of any of Examples 15-25, and wherein determining the node-specific adjustments comprises determining at least one of a processor throttle adjustment, a memory usage adjustment, a network bandwidth adjustment, or a fan speed adjustment.
Example 27 includes the subject matter of any of Examples 15-26, and wherein applying the determined adjustments comprises issuing a request to perform a live migration of a workload between the managed nodes.
Example 28 includes the subject matter of any of Examples 15-27, and wherein applying the determined adjustments comprises issuing a request to one of the managed nodes to apply one or more node-specific adjustments indicative of changes to an availability of one or more resources of the managed node to one or more of the workloads performed by the managed node.
Example 29 includes one or more machine-readable storage media comprising a plurality of instructions stored thereon that in response to being executed, cause an orchestrator server to perform the method of any of Examples 15-28.
Example 30 includes an orchestrator server to manage workloads among a plurality of managed nodes coupled to a network, the orchestrator server comprising one or more processors; communication circuitry coupled to the one or more processors; one or more memory devices having stored therein a plurality of instructions that, when executed by the one or more processors, cause the orchestrator server to perform the method of any of Examples 15-28.
Example 31 includes an orchestrator server to utilize availability data for a set of managed nodes to assign workloads, the orchestrator server comprising resource manager circuitry to assign workloads to the managed nodes; telemetry monitor circuitry to receive availability data from the managed nodes, wherein the availability data is indicative of a determination by each of the managed nodes as to an availability of the managed node to receive an additional workload, and receive telemetry data from the managed nodes, wherein the telemetry data is indicative of resource utilization by each of the managed nodes as the workloads are performed; wherein the resource manager circuitry is further to determine, as a function of the availability data, a reduced set of available managed nodes for analysis, determine, as a function of the telemetry data, adjustments to the workload assignments to increase the resource utilization among the reduced set of managed nodes, and apply the determined adjustments to the reduced set of managed nodes as the workloads are performed.
Example 32 includes the subject matter of Example 31, and wherein to assign the workloads comprises to assign a priority to one or more of the workloads.
Example 33 includes the subject matter of any of Examples 31 and 32, and wherein to assign a priority to one or more of the workloads comprises to assign a deterministic execution priority to one or more of the workloads.
Example 34 includes the subject matter of any of Examples 31-33, and wherein to assign the workloads comprises to generate availability data as a function of the assignment of the workloads.
Example 35 includes the subject matter of any of Examples 31-34, and wherein to determine, as a function of the telemetry data, adjustments to the workload assignments comprises to generate, as a function of the telemetry data, data analytics as the workloads are performed.
Example 36 includes the subject matter of any of Examples 31-35, and wherein to generate data analytics comprises to limit the generation of the data analytics to the reduced set of managed nodes.
Example 37 includes the subject matter of any of Examples 31-36, and wherein to generate data analytics comprises to identify trends in resource utilization of the workloads performed by the managed nodes in the reduced set of managed nodes.
Example 38 includes the subject matter of any of Examples 31-37, and wherein to generate data analytics comprises to generate profiles of the workloads performed by the managed nodes in the reduced set of managed nodes.
Example 39 includes the subject matter of any of Examples 31-38, and wherein to generate data analytics comprises to predict future resource utilization of the workloads performed by the managed nodes in the reduced set of managed nodes.
Example 40 includes the subject matter of any of Examples 31-39, and further including policy manager circuitry to obtain policy data indicative of one or more goals to be achieved in the management of the workloads, wherein the resource manager circuitry is further to modify the adjustments as a function of the policy data.
Example 41 includes the subject matter of any of Examples 31-40, and wherein to determine the adjustments comprises to determine one or more node-specific adjustments indicative of changes to an availability of one or more resources of a managed node in the reduced set of managed nodes to one or more of the workloads performed by the managed node.
Example 42 includes the subject matter of any of Examples 31-41, and wherein to determine the node-specific adjustments comprises to determine at least one of a processor throttle adjustment, a memory usage adjustment, a network bandwidth adjustment, or a fan speed adjustment.
Example 43 includes the subject matter of any of Examples 31-42, and wherein to apply the determined adjustments comprises to issue a request to perform a live migration of a workload between the managed nodes.
Example 44 includes the subject matter of any of Examples 31-43, and wherein to apply the determined adjustments comprises to issue a request to one of the managed nodes to apply one or more node-specific adjustments indicative of changes to an availability of one or more resources of the managed node to one or more of the workloads performed by the managed node.
Example 45 includes an orchestrator server to manage workloads among a plurality of managed nodes coupled to a network, the orchestrator server comprising circuitry for assigning workloads managed nodes; circuitry for receiving availability data from the managed nodes, wherein the availability data is indicative of a determination by each of the managed nodes as to an availability of the managed node to receive an additional workload; circuitry for receiving telemetry data from the managed nodes, wherein the telemetry data is indicative of resource utilization by each of the managed nodes as the workloads are performed; means for determining, as a function of the availability data, a reduced set of available managed nodes for analysis; means for determining, as a function of the telemetry data, adjustments to the workload assignments to increase the resource utilization among the reduced set of managed nodes; and means for applying the determined adjustments to the reduced set of managed nodes as the workloads are performed.
Example 46 includes the subject matter of Example 45, and wherein the circuitry for assigning the workloads comprises circuitry for assigning a priority to one or more of the workloads.
Example 47 includes the subject matter of any of Examples 45 and 46, and wherein the circuitry for assigning a priority to one or more of the workloads comprises to assign a deterministic execution priority to one or more of the workloads.
Example 48 includes the subject matter of any of Examples 45-47, and wherein the circuitry for assigning the workloads comprises circuitry for generating availability data as a function of the assignment of the workloads.
Example 49 includes the subject matter of any of Examples 45-48, and wherein the means for determining, as a function of the telemetry data, adjustments to the workload assignments comprises means for generating, as a function of the telemetry data, data analytics as the workloads are performed.
Example 50 includes the subject matter of any of Examples 45-49, and wherein the means for generating data analytics comprises means for limiting the generation of the data analytics to the reduced set of managed nodes.
Example 51 includes the subject matter of any of Examples 45-50, and wherein the means for generating data analytics comprises means for identifying trends in resource utilization of the workloads performed by the managed nodes in the reduced set of managed nodes.
Example 52 includes the subject matter of any of Examples 45-51, and wherein the means for generating data analytics comprises means for generating profiles of the workloads performed by the managed nodes in the reduced set of managed nodes.
Example 53 includes the subject matter of any of Examples 45-52, and wherein the means for generating data analytics means for predicting future resource utilization of the workloads performed by the managed nodes in the reduced set of managed nodes.
Example 54 includes the subject matter of any of Examples 45-53, and further including circuitry for obtaining policy data indicative of one or more goals to be achieved in the management of the workloads; and means for modifying the adjustments as a function of the policy data.
Example 55 includes the subject matter of any of Examples 45-54, and wherein the means for determining the adjustments comprises means for determining one or more node-specific adjustments indicative of changes to an availability of one or more resources of a managed node in the reduced set of managed nodes to one or more of the workloads performed by the managed node.
Example 56 includes the subject matter of any of Examples 45-55, and wherein the means for determining the node-specific adjustments comprises means for determining at least one of a processor throttle adjustment, a memory usage adjustment, a network bandwidth adjustment, or a fan speed adjustment.
Example 57 includes the subject matter of any of Examples 45-56, and wherein the means for applying the determined adjustments comprises means for issuing a request to perform a live migration of a workload between the managed nodes.
Example 58 includes the subject matter of any of Examples 45-57, and wherein the means for applying the determined adjustments comprises means for issuing a request to one of the managed nodes to apply one or more node-specific adjustments indicative of changes to an availability of one or more resources of the managed node to one or more of the workloads performed by the managed node.
Example 59 includes a managed node for providing availability data to an orchestrator server, the managed node comprising one or more processors; communication circuitry coupled to the one or more processors; one or more memory devices having stored therein a plurality of instructions that, when executed by the one or more processors, cause the managed node to receive a workload from the orchestrator server; generate telemetry data indicative of resource utilization as the workload is performed; compare the telemetry data to one or more predefined thresholds to provide availability data indicative of an availability of the managed node to receive an additional workload; report the availability data to be used by the orchestrator server to adjust workload assignments.
Example 60 includes the subject matter of Example 59, and wherein to report the availability data comprises to report the availability data based on a foraging algorithm.
Example 61 includes the subject matter of any of Examples 59 and 60, and wherein the plurality of instructions, when executed by the one or more processors, further cause the managed node to receive an indication of a priority of the workload from the orchestrator server; and wherein to compare the telemetry data to the one or more predefined thresholds comprises to select at least one predefined threshold as a function of the priority of the workload.
Example 62 includes the subject matter of any of Examples 59-61, and wherein to receive an indication of a priority comprises to receive an indication that the workload is to be executed deterministically; and to select at least one predefined threshold comprises to select at least one threshold associated with deterministic execution.
Example 63 includes the subject matter of any of Examples 59-62, and wherein to compare the telemetry data to the one or more predefined thresholds comprises to compare processor utilization data to a processor availability threshold.
Example 64 includes the subject matter of any of Examples 59-63, and wherein to compare the telemetry data to the one or more predefined thresholds comprises to compare memory utilization data to a memory availability threshold.
Example 65 includes the subject matter of any of Examples 59-64, and wherein to report the availability data comprises to report the availability data to another managed node to be reported to the orchestrator server.
Example 66 includes the subject matter of any of Examples 59-65, and wherein to report the availability data comprises to report the availability data directly to the orchestrator server.
Example 67 includes the subject matter of any of Examples 59-66, and wherein the plurality of instructions, when executed by the one or more processors, further cause the managed node to receive additional availability data from at least one other managed node; and to report the availability data comprises to report the generated availability data and the additional availability data to the orchestrator server.
Example 68 includes the subject matter of any of Examples 59-67, and wherein to receive additional availability data from at least one other managed node comprises to receive additional availability data from at least one other managed node with a predefined relationship to the managed node.
Example 69 includes the subject matter of any of Examples 59-68, and wherein to receive additional availability data from at least one other managed node comprises to receive additional availability data from at least one other managed node identified in a predefined set of managed nodes.
Example 70 includes the subject matter of any of Examples 59-69, and wherein to receive additional availability data from at least one other managed node comprises to receive additional availability data from at least one other managed node within a predefined proximity of the managed node.
Example 71 includes the subject matter of any of Examples 59-70, and wherein the plurality of instructions, when executed by the one or more processors, further cause the managed node to receive a request for the availability data from the orchestrator server; and wherein to report the availability data comprises to report, in response to the request, the availability data.
Example 72 includes the subject matter of any of Examples 59-71, and wherein the plurality of instructions, when executed by the one or more processors, further cause the managed node to receive a request for the availability data from another managed node; and wherein to report the availability data comprises to report, in response to the request, the availability data.
Example 73 includes the subject matter of any of Examples 59-72, and wherein the plurality of instructions, when executed by the one or more processors, further cause the managed node to receive node-specific adjustments from the orchestrator server, wherein the node-specific adjustments are indicative of at least one of a processor throttle adjustment, a memory usage adjustment, a network bandwidth adjustment, or a fan speed adjustment; and execute the workload with the node-specific adjustments.
Example 74 includes a method for providing availability data to an orchestrator server, the method comprising receiving, by a managed node, a workload from the orchestrator server; generating, by the managed node, telemetry data indicative of resource utilization as the workload is performed; comparing, by the managed node, the telemetry data to one or more predefined thresholds to provide availability data indicative of an availability of the managed node to receive an additional workload; and reporting, by the managed node, the availability data to be used by the orchestrator server to adjust workload assignments.
Example 75 includes the subject matter of Example 74, and wherein reporting the availability data comprises reporting the availability data based on a foraging algorithm
Example 76 includes the subject matter of any of Examples 74 and 75, and further including receiving, by the managed node, an indication of a priority of the workload from the orchestrator server; and wherein comparing the telemetry data to the one or more predefined thresholds comprises selecting at least one predefined threshold as a function of the priority of the workload.
Example 77 includes the subject matter of any of Examples 74-76, and wherein receiving an indication of a priority comprises receiving an indication that the workload is to be executed deterministically; and selecting at least one predefined threshold comprises to select at least one threshold associated with deterministic execution.
Example 78 includes the subject matter of any of Examples 74-77, and wherein comparing the telemetry data to the one or more predefined thresholds comprises comparing processor utilization data to a processor availability threshold.
Example 79 includes the subject matter of any of Examples 74-78, and wherein comparing the telemetry data to the one or more predefined thresholds comprises comparing memory utilization data to a memory availability threshold.
Example 80 includes the subject matter of any of Examples 74-79, and wherein reporting the availability data comprises reporting the availability data to another managed node to be reported to the orchestrator server.
Example 81 includes the subject matter of any of Examples 74-80, and wherein reporting the availability data comprises reporting the availability data directly to the orchestrator server.
Example 82 includes the subject matter of any of Examples 74-81, and further including receiving, by the managed node, additional availability data from at least one other managed node; and reporting the availability data comprises reporting the generated availability data and the additional availability data to the orchestrator server.
Example 83 includes the subject matter of any of Examples 74-82, and wherein receiving additional availability data from at least one other managed node comprises receiving additional availability data from at least one other managed node with a predefined relationship to the managed node.
Example 84 includes the subject matter of any of Examples 74-83, and wherein receiving additional availability data from at least one other managed node comprises receiving additional availability data from at least one other managed node identified in a predefined set of managed nodes.
Example 85 includes the subject matter of any of Examples 74-84, and wherein receiving additional availability data from at least one other managed node comprises receiving additional availability data from at least one other managed node within a predefined proximity of the managed node.
Example 86 includes the subject matter of any of Examples 74-85, and further including receiving, by the managed node, a request for the availability data from the orchestrator server; and wherein reporting the availability data comprises reporting, in response to the request, the availability data.
Example 87 includes the subject matter of any of Examples 74-86, and further including receiving, by the managed node, a request for the availability data from another managed node; and wherein reporting the availability data comprises reporting, in response to the request, the availability data.
Example 88 includes the subject matter of any of Examples 74-87, and further including receiving, by the managed node, node-specific adjustments from the orchestrator server, wherein the node-specific adjustments are indicative of at least one of a processor throttle adjustment, a memory usage adjustment, a network bandwidth adjustment, or a fan speed adjustment; and executing, by the managed node, the workload with the node-specific adjustments.
Example 89 includes one or more machine-readable storage media comprising a plurality of instructions stored thereon that in response to being executed, cause a managed node to perform the method of any of Examples 59-88.
Example 90 includes a managed node for providing availability data to an orchestrator server, the managed node comprising one or more processors; communication circuitry coupled to the one or more processors; one or more memory devices having stored therein a plurality of instructions that, when executed by the one or more processors, cause the managed node to perform the method of any of Examples 59-88.
Example 91 includes a managed node for providing availability data to an orchestrator server, the managed node comprising workload executor circuitry to receive a workload from the orchestrator server; telemetry data generator circuitry to generate telemetry data indicative of resource utilization as the workload is performed; and availability data manager circuitry to compare the telemetry data to one or more predefined thresholds to provide availability data indicative of an availability of the managed node to receive an additional workload, and report the availability data to be used by the orchestrator server to adjust workload assignments.
Example 92 includes the subject matter of Example 91, and wherein to report the availability data comprises to report the availability data based on a foraging algorithm.
Example 93 includes the subject matter of any of Examples 91 and 92, and wherein the workload executor circuitry is further to receive an indication of a priority of the workload from the orchestrator server, and wherein to compare the telemetry data to the one or more predefined thresholds comprises to select at least one predefined threshold as a function of the priority of the workload.
Example 94 includes the subject matter of any of Examples 91-93, and wherein to receive an indication of a priority comprises to receive an indication that the workload is to be executed deterministically; and to select at least one predefined threshold comprises to select at least one threshold associated with deterministic execution.
Example 95 includes the subject matter of any of Examples 91-94, and wherein to compare the telemetry data to the one or more predefined thresholds comprises to compare processor utilization data to a processor availability threshold.
Example 96 includes the subject matter of any of Examples 91-95, and wherein to compare the telemetry data to the one or more predefined thresholds comprises to compare memory utilization data to a memory availability threshold.
Example 97 includes the subject matter of any of Examples 91-96, and wherein to report the availability data comprises to report the availability data to another managed node to be reported to the orchestrator server.
Example 98 includes the subject matter of any of Examples 91-97, and wherein to report the availability data comprises to report the availability data directly to the orchestrator server.
Example 99 includes the subject matter of any of Examples 91-98, and wherein the availability data manager is further to receive additional availability data from at least one other managed node, and wherein to report the availability data comprises to report the generated availability data and the additional availability data to the orchestrator server.
Example 100 includes the subject matter of any of Examples 91-99, and wherein to receive additional availability data from at least one other managed node comprises to receive additional availability data from at least one other managed node with a predefined relationship to the managed node.
Example 101 includes the subject matter of any of Examples 91-100, and wherein to receive additional availability data from at least one other managed node comprises to receive additional availability data from at least one other managed node identified in a predefined set of managed nodes.
Example 102 includes the subject matter of any of Examples 91-101, and wherein to receive additional availability data from at least one other managed node comprises to receive additional availability data from at least one other managed node within a predefined proximity of the managed node.
Example 103 includes the subject matter of any of Examples 91-102, and wherein the availability data manager is further to receive a request for the availability data from the orchestrator server, and wherein to report the availability data comprises to report, in response to the request, the availability data.
Example 104 includes the subject matter of any of Examples 91-103, and wherein the availability data manager circuitry is further to receive a request for the availability data from another managed node, and wherein to report the availability data comprises to report, in response to the request, the availability data.
Example 105 includes the subject matter of any of Examples 91-104, and wherein the workload executor circuitry is further to receive node-specific adjustments from the orchestrator server, wherein the node-specific adjustments are indicative of at least one of a processor throttle adjustment, a memory usage adjustment, a network bandwidth adjustment, or a fan speed adjustment; and execute the workload with the node-specific adjustments.
Example 106 includes a managed node for providing availability data to an orchestrator server, the managed node comprising circuitry for receiving a workload from the orchestrator server; means for generating telemetry data indicative of resource utilization as the workload is performed; means for comparing the telemetry data to one or more predefined thresholds to provide availability data indicative of an availability of the managed node to receive an additional workload; means for reporting the availability data to be used by the orchestrator server to adjust workload assignments.
Example 107 includes the subject matter of Example 106, and wherein the means for reporting the availability data comprises means for reporting the availability data based on a foraging algorithm
Example 108 includes the subject matter of any of Examples 106 and 107, and further including circuitry for receiving an indication of a priority of the workload from the orchestrator server; and wherein the means for comparing the telemetry data to the one or more predefined thresholds comprises means for selecting at least one predefined threshold as a function of the priority of the workload.
Example 109 includes the subject matter of any of Examples 106-108, and wherein the circuitry for receiving an indication of a priority comprises circuitry for receiving an indication that the workload is to be executed deterministically; and wherein the means for selecting at least one predefined threshold comprises means for selecting at least one threshold associated with deterministic execution.
Example 110 includes the subject matter of any of Examples 106-109, and wherein the means for comparing the telemetry data to the one or more predefined thresholds comprises means for comparing processor utilization data to a processor availability threshold.
Example 111 includes the subject matter of any of Examples 106-110, and wherein the means for compare the telemetry data to the one or more predefined thresholds comprises means for comparing memory utilization data to a memory availability threshold.
Example 112 includes the subject matter of any of Examples 106-111, and wherein the means for reporting the availability data comprises means for reporting the availability data to another managed node to be reported to the orchestrator server.
Example 113 includes the subject matter of any of Examples 106-112, and wherein the means for reporting the availability data comprises means for reporting the availability data directly to the orchestrator server.
Example 114 includes the subject matter of any of Examples 106-113, and further including circuitry for receiving additional availability data from at least one other managed node; and the means for reporting the availability data comprises means for reporting the generated availability data and the additional availability data to the orchestrator server.
Example 115 includes the subject matter of any of Examples 106-114, and wherein the circuitry for receiving additional availability data from at least one other managed node comprises circuitry for receiving additional availability data from at least one other managed node with a predefined relationship to the managed node.
Example 116 includes the subject matter of any of Examples 106-115, and wherein the circuitry for receiving additional availability data from at least one other managed node comprises circuitry for receiving additional availability data from at least one other managed node identified in a predefined set of managed nodes.
Example 117 includes the subject matter of any of Examples 106-116, and wherein the circuitry for receiving additional availability data from at least one other managed node comprises circuitry for receiving additional availability data from at least one other managed node within a predefined proximity of the managed node.
Example 118 includes the subject matter of any of Examples 106-117, and further including circuitry for receiving a request for the availability data from the orchestrator server; and wherein the means for reporting the availability data comprises means for reporting, in response to the request, the availability data.
Example 119 includes the subject matter of any of Examples 106-118, and further including circuitry for receiving a request for the availability data from another managed node; and wherein the means for reporting the availability data comprises means for reporting, in response to the request, the availability data.
Example 120 includes the subject matter of any of Examples 106-119, and further including circuitry for receiving node-specific adjustments from the orchestrator server, wherein the node-specific adjustments are indicative of at least one of a processor throttle adjustment, a memory usage adjustment, a network bandwidth adjustment, or a fan speed adjustment; and means for executing the workload with the node-specific adjustments.
The present application claims the benefit of U.S. Provisional Patent Application No. 62/365,969, filed Jul. 22, 2016, U.S. Provisional Patent Application No. 62/376859, filed Aug. 18, 2016, and U.S. Provisional Patent Application No. 62/427,268, filed Nov. 29, 2016.
Number | Date | Country | |
---|---|---|---|
62365969 | Jul 2016 | US | |
62376859 | Aug 2016 | US | |
62427268 | Nov 2016 | US |