Abstract: Embodiments of the present disclosure provide a computer-implemented method and an apparatus for a storage system. The method comprises: in response to receiving a read request of a first container for data in a storage device, obtaining an identifier associated with the read request; searching for metadata of the read request in a metadata set based on the identifier, the metadata recording addressing information of the read request, the metadata set including metadata of access requests for the storage device during a past period; and in response to finding the metadata of the read request in the metadata set, determining, based on the metadata, a cached page of a second container storing the data; and providing the cached page from the second container to the first container to avoid reading the data from the storage device.

Abstract: An embodiment of the present invention evaluates plural expressions. A model is generated and configured to evaluate a plurality of expressions each including one or more expression tokens and indicating a data pattern. The model includes a plurality of nodes with one or more of the nodes associated with an expression token and one or more links between the nodes. The links are associated with information indicating each expression including each expression token associated with nodes connected by the links. Data including one or more data tokens is applied to the model. The nodes of the model are traversed over one or more corresponding links based on the one or more data tokens within the data corresponding to expression tokens associated with the nodes. Expressions corresponding to the data are determined based on the expressions associated with the one or more corresponding links.

Abstract: An intelligent placement engine generates a placement map that provides a configuration for deploying a service based at least in part, on one or more configuration parameters. A data center in which the service is to be hosted is defined using a data center definition, while the service is defined using a service definition. The configuration parameters include estimated probabilities calculated based on estimated resource consumption data. The resource consumption data is estimated based at least in part on historical data distributions.

Abstract: A method, in a data processing system comprising a processor and a memory, for managing computing tasks, the method comprising retrieving, by the data processing system, data files from a corpus data server, identifying, by the data processing system, task features from the data files, creating, by the data processing system, a plurality of tasks for loading to a memory device and execution by one or more processors based on the task features, associating, by the data processing system, the plurality of tasks based on similarities of the task features, and generating, by the data processing system, mappings of the plurality of tasks based on the association, wherein the mappings include instructions for logically merging execution of the plurality of tasks.

Abstract: An accelerator manager monitors and logs performance of multiple accelerators, analyzes the logged performance, determines from the logged performance of a selected accelerator a desired programmable device for the selected accelerator, and specifies the desired programmable device to one or more accelerator developers. The accelerator manager can further analyze the logged performance of the accelerators, and generate from the analyzed logged performance an ordered list of test cases, ordered from fastest to slowest. A test case is selected, and when the estimated simulation time for the selected test case is less than the estimated synthesis time for the test case, the test case is simulated and run. When the estimated simulation time for the selected test case is greater than the estimated synthesis time for the text case, the selected test case is synthesized and run.

Abstract: A method for resource allocation and a terminal device are provided. The method includes the following. An operating system first identifies an internal running scenario of a target application running in a foreground of a terminal device. The operating system then acquires a plurality of resource adjustment strategies associated with preset system resources to-be-adjusted in the internal running scenario. The operating system acquires status information of the internal running scenario. Thereafter, the operating system adjusts usage priorities of the plurality of resource adjustment strategies according to the status information. The operating system adjusts the preset system resources to-be-adjusted by selecting a resource adjustment strategy with the highest priority.

Abstract: Systems, methods, apparatus, and a computer-readable medium are described for generating and receiving information regarding the current state of the active virtual machine in the host and switching the standby virtual machine from standby to active in response to determining that the current active virtual machine may no longer be able to service network packets.

Abstract: A method may include deploying a plurality of container pods to a plurality of container nodes in a container environment. Each of the plurality of container pods may include one or more services. Each of the plurality of container nodes may include one or more container pods. The plurality of container pods may be deployed to the plurality of container nodes based on initial characterizations of usage factors for each of the plurality of container pods. The method may also include monitoring actual usage factors for each of the plurality of container pods after deployment to the plurality of container nodes; identifying one or more container pods in the plurality of container pods that deviate from their initial characterizations of usage factors; and redistributing the one or more container pods throughout the plurality of container nodes based on the actual usage factors.

Abstract: A system for data communication. In one example method, a first system sends an enhanced capability exchange (CAPEX) request message to a second system. The CAPEX request message includes a request to change the number of connection pipes on an established socket-based connection between the first system and the second system. The first system receives an enhanced CAPEX response message from the second system. The CAPEX response message accepts the request to change the number of connection pipes on the established connection. The first system changes the number of connection pipes on the established connection in accordance with the accepted request.

Abstract: System and method for managing storage requests issued from multiple sources in a distributed storage system utilizes different queues at a host computer in the distributed storage system to place different classes of storage requests for access to a virtual storage area network. The storage requests in the queues are processed using a fair scheduling algorithm. For each queue, when the storage requests in the queue exceeds a threshold, a backpressure signal is generated and transmitted to at least one source for a class of storage requests queued in one of the queues corresponding to that backpressure signal to delay issuance of new storage requests of that class of storage requests.

Abstract: Systems and methods are described for bootstrapping an application with metadata specified in a template. The template specifies a stack of resources that will be used to execute an application and also includes a set of metadata for customizing the resources and the application. When the system receives the template, it instantiates a compute node which will execute the application. The compute node may contain at least one initialization script for bootstrapping the application with the metadata contained in the template. This functionality allows users to bootstrap the application running on the compute node with data from within the template that was used to create the stack of resources for executing the application. In this manner, metadata, configuration files, package names and versions can be passed by the application owner to the remote compute node.

Abstract: Conversations are modeled using dependency graph data structures to facilitate the interaction of users with automated assistants when performing actions performed by computing services. An automated assistant may utilize a dependency graph data structure to guide or otherwise control a human-to-computer dialog session with a user, e.g., by generating one or more outputs or prompts that are presented to the user on a computing device operated by that user, and may thereby enable efficient use of technical hardware.

Abstract: Various examples are disclosed for predictive allocation of computing resources in a virtual desktop infrastructure environment. A computing environment can generate a predictive usage model as a time series that forecasts a usage of virtual machines that provide virtual desktop sessions in a virtual desktop infrastructure environment. The computing environment can determine a number of the plurality of virtual machines that will be operating at a future time utilizing the predictive usage model and identify computing resources required for the number of the virtual machines to operate at the future time. The computing environment can allocate the computing resources such that the computing resources are available at the future time.

Abstract: A method to deploy a multitier application in a virtualized computing environment includes receiving an open virtualization format (OVF) package. The OVF package includes an OVF descriptor, one or more virtual disk image files of virtual machines, and a multitier application blueprint specifying software components on the virtual machines and dependencies of the software components. The method further includes deploying the virtual machines based on the OVF package and executing the multitier application blueprint by deploying the software components on the virtual machines pursuant to the dependencies of the software components.

Abstract: Provided are an infrastructure apparatus and method of providing collaboration between thing devices. Since a plurality of thing devices located in a specific space share an experience where the plurality of thing devices have performed a collaboration service, each of the thing devices may search for peripheral thing devices for performing the collaboration service without intervention of a center server and may generate a collaboration group including the found thing devices, and tasks for performing through negotiation may be autonomously distributed to the thing devices included in the collaboration group, thereby providing autonomous collaboration between the thing devices.

Abstract: A system for controlling power to a complex system, comprising a plurality of processors, one or more power supply unit, each power supply unit including an overcurrent warning system and a baseboard management controller coupled to the plurality of processors and the one or more power supply unit, wherein the baseboard management controller is configured to determine whether a power cap control has been enabled and to adjust an overcurrent warning threshold of the overcurrent warning system.

Abstract: A distributed application is hosted on plural virtual machines running on a public cloud. The virtual machines are placed “permanently” in a diagnostic mode, during which diagnostic data is captured by the public cloud for each virtual machine for each capture period, e.g., each minute. Captured diagnostic data is “retired”, that is, deleted, after a retirement period, e.g., 10 days. A cloud management service extracts the diagnostic data periodically from the public cloud using to an extraction period comparable to the capture period. The extracted data is aggregated across virtual machines to obtain group utilization time profiles for the distributed application. A configuration of the distributed application is then modified at least in part based on the group utilization time profiles.

Abstract: Disclosed herein are an apparatus and method for providing long-term function execution in a serverless environment. The method for providing long-term function execution in a serverless environment is performed by an apparatus for providing long-term function execution in a serverless environment, and includes registering a long-term function execution proxy when a long-term execution request is received from a client, allocating a long-term function executor corresponding to the long-term execution request, executing, by the long-term function execution proxy, a long-term function using the allocated long-term function executor, and storing execution results of the long-term function.

Abstract: In one embodiment, a system for managing communication connections in a virtualization environment includes (1) a first host machine implementing a virtualization environment based on a first platform, wherein the first host machine includes a first hypervisor, at least one virtual machine, and one or more virtual infrastructure elements and (2) a virtual disk including a plurality of storage devices. A management module for the system may perform steps to convert the first host machine to a second platform by installing (on the first host machine) a second hypervisor associated with the second platform, disabling the first hypervisor, capturing a configuration describing elements of a virtual infrastructure associated with the first hypervisor, registering the captured configuration with the second hypervisor, creating elements of the captured configuration in the context of the second platform and in association with the second hypervisor, and then enabling the second hypervisor.

Abstract: Computational methods and systems that proactively manage usage of computational resources of a distributed computing system are described. A sequence of metric data representing usage of a resource is detrended to obtain a sequence of non-trendy metric data. Stochastic process models, a pulse wave model and a seasonal model of the sequence of non-trendy metric data are computed. When a forecast request is received, a sequence of forecasted metric data is computed over a forecast interval based on the estimated trend and one of the pulse wave or seasonal model that matches the periodicity of the sequence of non-trendy metric data. Alternatively, the sequence of forecasted metric data is computed based on the estimated trend and the stochastic process model with a smallest accumulated residual error. Usage of the resource by virtual objects of the distributed computing system may be adjusted based on the sequence of forecasted metric data.

Abstract: A software service execution apparatus comprising a registry of software services, each service to execute a data processing function in the registry; a controller, to receive a processing request defining requested data processing functions, compose an execution schedule of software services from the software services identified, fulfill the data processing request by identifying a software service that matches the requested data processing function, include the identified software service in the execution schedule, and control execution of the schedule. The apparatus further comprises a machine learning mechanism configured to maintain a record of the composing.

Abstract: A method and system for optimizing the interaction and execution of multiple service tasks associated with a logical transaction. The multiple components or “legs” of the transaction consisting of tasks executable by a computing service or software as a service (SAAS) endpoint are identified. The system determines a strategy type or belief level associated with each of the service tasks included in a transaction. The belief level may be categorized as either “optimistic” or “pessimistic” based on one or more performance parameters (e.g., a probability of failure of a service and an expense associated with a failure of the service) derived from historical data associated with a particular transaction or service task. A sequence of execution for the multiple service tasks associated with the transaction is determined based at least in part on the belief level associated with each of the multiple service tasks.

Abstract: A method, apparatus, and computer program product for synchronously starting programs on multiple devices connected to a server is provided. A synchronous point of a program to be synchronously started for each of the multiple devices is identified. A wait function is dynamically injected into the synchronous point for each of the multiple devices. A start time from the server is received in response to the multiple devices entering a waiting state. The programs are synchronously started in response to the start time arriving for each of the multiple devices.

Abstract: Various avionics systems may benefit from providing remote L-band smart antennas. For example, smart antennas may be used in connection with transponders and other avionics equipment. The smart antenna may incorporate functionality that may obviate the need for coaxial RF cabling. According to certain embodiments, an avionics system may include an antenna having at least a radio frequency function included therewith. The avionics system may also include a data bus connecting the antenna to a digital receiver at avionics processing hardware.

Abstract: The present disclosure provides for computational data analysis across multiple data sources. A pipeline (or workflow) is imported and a dataset is selected. The dataset resides on a virtual file system and includes data residing on one or more storage locations associated with the virtual file system. One or more compute resources are selected to perform the pipeline analysis based at least on the imported pipeline and the dataset. The one or more compute resources are selected from a plurality of available compute resources associated with the one or more storage locations associated with the virtual file system. The pipeline analysis is performed using the selected compute resources on the dataset in one or more secure clusters. The resulting data generated from the pipeline analysis is submitted to the virtual file system.

Abstract: Technologies for dynamically allocating resources among a set of managed nodes include an orchestrator server to receive telemetry data from the managed nodes indicative of resource utilization and workload performance by the managed nodes as the workloads are executed, generate a resource allocation map indicative of allocations of resources among the managed nodes, determine, as a function of the telemetry data and the resource allocation map, a dynamic adjustment to allocation of resources to at least one of the managed nodes to improve performance of at least one of the workloads executed on the at least one of the managed nodes, and apply the adjustment to the allocation of the resources among the managed nodes as the workloads are executed. Other embodiments are also described and claimed.

Abstract: An apparatus includes a programmable circuit that configures circuits for executing tasks. The apparatus estimates an execution time-period required for executing a first task by first circuits configured in the programmable circuit, and determines a configuration number indicating a number of second circuits that are to be configured, in the programmable circuit, for executing a second task to be executed after the first task, based on the execution time-period and a configuration time-period required for configuring the configuration number of the second circuits in the programmable circuit. The apparatus causes the programmable circuit to configure, during execution of the first task, the configuration number of the second circuits, and adjusts the configuration number, based on a relationship between a time at which the first task is completed and a time at which configuration of the configuration number of the second circuits in the programmable circuit is completed.

Abstract: Systems, apparatuses, and methods related to a block-based processor core topology register are disclosed. In one example of the disclosed technology, a processor can include a plurality of block-based processor cores for executing a program including a plurality of instruction blocks. A respective block-based processor core can include a sharable resource and a programmable composition topology register. The programmable composition topology register can be used to assign a group of the physical processor cores that share the sharable resource.

Abstract: Concepts and technologies are disclosed herein for providing network address translation in a software defined networking environment. A control system can detect a request to create a network address translation function. The control system can analyze a recipe to identify a network address translation switch and a network address translation virtual function that are to provide functionality of the network address translation function, trigger instantiation of the network address translation switch via a network control function, and trigger instantiation of the network address translation virtual function via a service control function. The control system can validate the network address translation function and chain the network address translation function to a host. The network address translation function can provide address translation for the host using the network address translation switch.

Abstract: A method for launching a plurality of computing instances may include obtaining a request to launch a plurality of computing instances. The request may indicate a first number and a target number of compute instances that is larger than the first number. The method may further include verifying that there is sufficient capacity, for example of a compute instance service, to launch at least the first number of compute instances. The request may be fulfilled by at least launching at least the first number of compute instances in a way that bypasses a rate limit that limits a rate at which compute instances can be launched. In some cases, the method may additionally include launching additional compute instances, as they become available, until the target number has been reached.

Abstract: A disclosed policy introduced effect prediction apparatus includes a memory storing a set of instructions of a policy introduced effect prediction program, and processors programed to execute the instructions to perform a policy introduced effect prediction process. The policy introduced effect prediction process includes calculating at least one calculation model based on first difference data calculated based on time series data of an objective variable and second difference data calculated based on time series data of an explanatory variable, calculating a difference amount of the objective variable by inputting a difference amount of the explanatory variable for introducing a policy into the calculated calculation model, and calculating a predicted value of the objective variable in response to introduction of the policy by adding the calculated difference amount of the objective variable to a predicted value of the objective variable in the absence of the introduction of the policy.

Abstract: A non-transitory computer-readable storage medium may comprise instructions stored thereon. When executed by at least one processor, the instructions may be configured to cause a computing device to implement at least a user interface module and a coordinator module. The user interface module may be configured to pass layout constraints to the coordinator module. The coordinator module may be configured to process input received by the computing device, recognize gestures based on the input, update a display based on the recognized gestures and the layout constraints passed to the coordinator module by the user interface module, and in response to recognizing an application-specific gesture, pass the application-specific gesture to the user interface module.

Abstract: Embodiments provide optimized deployment of workloads. A first workload to be deployed in at least one data center of a plurality of data centers is received. A first plurality of candidate data centers is selected from the plurality of data centers, and the first workload is deployed to each of the first plurality of candidate data centers. A first performance metric is collected from each respective data center of the first plurality of candidate data centers based on execution of the first workload deployed at the respective data centers, and a first optimal data center from the first plurality of candidate data centers is identified based on the first performance metrics. The first workload is removed from each of the first plurality of candidate data centers, other than the identified first optimal data center.

Abstract: A data analyzer engine receives feedback associated with delivery of content (such as streaming content) to multiple subscriber domains in a cable network environment. The data analyzer engine analyzes the feedback to identify attributes of current content consumption by playback devices operated in the subscriber domains. Based on the analyzed feedback, the data analyzer engine produces control information specifying data flow control rates in which the playback devices are assigned to retrieve content over a shared communication link in the cable network environment. Accordingly, data flow rates in a network environment (such as a subscriber domain) can be controlled to promote fairness and optimal use of available bandwidth amongst multiple playback devices sharing a respective communication link.

Abstract: Computing processes are allocated to a plurality of computing nodes while taking a precedence constraint among a plurality of computing processes into consideration even when a deadline is longer than a period of the computing process. In a distributed allocation device, a creation unit determines an order to execute the computing processes based on the precedence constraint, and creates a constraint condition equation satisfying the deadline based on an execution time of the computing processes and a data communication volume among the computing processes. If the deadline is longer than the entire period, the creation unit creates the constraint condition equation which does not allocate an other computing process during execution of an excess computing process, in the subsequent entire period, to the computing node to which the excess computing process is to be allocated, which is executed past the entire period.

Abstract: A method, system and computer program product for efficiently utilizing connections in connection pools. A period of time an application running on a virtual machine needs a greater number of connections to an external resource than allocated in its pool of connections is identified. The connection pool for this application as well as the connection pools for the other applications containing connections to the same external resource are merged to form a logical pool of connections to be shared by those applications during the identified period of time. Alternatively, in an application server cluster environment, the connection pools utilized by the application servers to access the external resource may be reconfigured based on the weight assigned to each member (or application server) of the cluster which is based on the member's load size. In these manners, the resource connections in these pools of connections will be more efficiently utilized.

Abstract: An example context-based computing framework includes methods to configure a computer device. Some example methods include generating a request for access to computing resources available at a first location in response to determining the computer device has moved from a second location to the first location. Some example methods also include, based on context data contained in a response to the request, configuring an interface of an application of the computer device is configured to allow interaction with a subset of a set of features of a first computing resource of the computing resources. The context data is identified in the context profile and the context profile is associated with the first location. Some example methods further include interacting, via the interface of the application, with a first feature of the subset of the set of features of the first computing resource.

Abstract: Concepts and technologies are disclosed herein for providing and using a deep packet inspection virtual function. A control system can detect a service request. The control system can analyze a policy to determine a function of a service to which the service request relates, a virtual machine that will host the function, and a deep packet inspection virtual function associated with the service. The control system can trigger loading of an image to the virtual machine and instantiation of the virtual machine. The image can include the function of the service and the deep packet inspection virtual function. The control system can validate the service and the deep packet inspection virtual function.

Abstract: A data processing system comprises one or more data processing units, a configurable interconnect and control circuitry. The control circuitry allocates one or more of the data processing units to a virtual machine and configures the configurable interconnect so as route one or more data processing tasks from the virtual machine to the one or more data processing units allocated for use by that virtual machine. This can provide a flexible and adaptable data processing system for carrying out the data processing tasks of a virtual machine, with the particular allocation of data processing units being substantially transparent to the virtual machine.

Abstract: A new processing architecture is described that utilizes a data processing unit (DPU). Unlike conventional compute models that are centered around a central processing unit (CPU), the DPU that is designed for a data-centric computing model in which the data processing tasks are centered around the DPU. The DPU may be viewed as a highly programmable, high-performance I/O and data-processing hub designed to aggregate and process network and storage I/O to and from other devices. The DPU comprises a network interface to connect to a network, one or more host interfaces to connect to one or more application processors or storage devices, and a multi-core processor with two or more processing cores executing a run-to-completion data plane operating system and one or more processing cores executing a multi-tasking control plane operating system. The data plane operating system is configured to support software functions for performing the data processing tasks.

Abstract: A hypervisor configures a page table entry in a host page table to map an address associated with memory-mapped input-output (MMIO) for a virtual device of a guest of the hypervisor to an input/output (I/O) instruction. The address is marked in the page table entry as a hypervisor exit entry, and the page table entry to cause an exit to the hypervisor responsive to the guest attempting to access the address. Responsive to detecting an exit to the hypervisor caused by the guest attempting to access the address, the hypervisor receives the I/O instruction mapped to the address that caused the exit. The hypervisor then executes the I/O instruction on behalf of the guest.

Abstract: As a web browser application executing on a computing renders a web application that contains a plurality of objects, the web browser application carries out, respectively for each object, a set of resource-consumption actions involving use of resources of the client device. Further, the web browser application dynamically manages operational states of the objects based on a cumulative resource-consumption of the web application, including (i) determining that the cumulative resource-consumption of the web application exceeds a defined limit, and (ii) responsive to the determining, transitioning at least one of the objects from a first operational state to a second operational state, where the object contributes less to the cumulative resource-consumption of the web application in the second operational state than in the first operational state.

Abstract: Embodiments include a technique for predictive virtual server scheduling and optimization of dynamic consumable resources to achieve priority-based workload performance objectives is provided. Techniques include calculating a capacity requirement for each virtual server based on an administrator defined performance objective for each virtual server. Techniques also include calculating an available CPU capacity for each importance level based on the calculated capacity requirement for each virtual server, sending the calculated available CPU capacity to a scheduler; and selecting a target hypervisor based on the calculated capacity requirement of each virtual server and the calculated available CPU capacity at each importance level of each hypervisor.

Abstract: Embodiments of the invention provide a method for load balancing a networking system comprising multiple computing nodes. The method comprises maintaining one or more data sets on at least one computing node. The method further comprises receiving, from each computing node, a load information unit for the computing node, wherein the load information unit relates to resource usage on the computing node. For each computing node, the method determines whether the load information for the computing node exceeds a corresponding load threshold for the computing node. A data set on at least one computing node is transferred to another computing node when the load information for the at least one computing node exceeds a corresponding load threshold for the at least one computing node.

Abstract: A method for protecting a device from information attacks, in which received from the at least one device are checked by a security device, included in the at least one device, for an information attack, and for the case that the received data are associated with an information attack, characteristic data concerning the information attack are stored in the at least one device and transmitted via a communication interface to a unit in communicative connection with a plurality of devices, and the unit evaluates the characteristic data transmitted from the at least one device and outputs a warning message to at least a portion of the devices in communicative connection with the unit as a function of at least one predefined criterion.

Abstract: A method includes selecting a given workload associated with at least one application type, and analyzing the given workload to determine a set of functional patterns describing resource structures for implementing functionality of the at least one application type. The method also includes determining resource consumption demand profiles for each of the set of functional patterns and determining micro workload distributions for each of the resource consumption demand profiles, a given one of the micro workload distributions comprising a number of micro workloads, each micro workload comprising a normalized unit of resource consumption metrics. The method further includes converting the micro workload distributions for each of the resource consumption demand profiles into a set of resource requirements for running the given workload on an information technology infrastructure, and allocating resources of the information technology infrastructure to the given workload based on the set of resource requirements.

Abstract: A method for mediating resource allocation includes a step of receiving from at least one first computation module, a first proposed resource allocation for completing a computational task. The method further includes a step of comparing the received first proposed resource allocation to at least one of a predetermined resource allocation criteria and at least one second proposed resource allocation received from at least one second computation module, and a step of determining a resource allocation instruction based, at least in part, on a result of the step of comparing.

Abstract: An apparatus in one embodiment comprises a processing platform configured to implement multi-layer infrastructure comprising compute, storage and network resources at a relatively low level of the multi-layer infrastructure, an application layer at a relatively high level of the multi-layer infrastructure, and one or more additional layers arranged between the relatively high level and the relatively low level. The processing platform is further configured to determine policies for respective different ones of the layers of the multi-layer infrastructure, the policy for a given one of the layers of the multi-layer infrastructure defining rules and requirements relating to that layer, to enforce the policies at the respective layers of the multi-layer infrastructure, and to monitor performance of an application executing in the multi-layer infrastructure. One or more configuration parameters of the multi-layer infrastructure are adjusted based at least in part on a result of the monitoring.

Abstract: A database system comprises a plurality of servers with data storage. The system holds first and second database tables. Records of the first database table are distributed among the servers according to values of a first distribution column. The servers can receive access requests and log workload data representative of the requests. Based on the workload data, the servers can determine a performance impact associated with columns of the tables and, upon determining that a performance impact associated with one of the columns is greater than a performance impact associated with the first distribution column, the servers can re-distribute the records among the plurality of servers according to values of the second column.

Abstract: Methods, apparatus, systems and articles of manufacture are disclosed for managing compute resources in a hyperconverged infrastructure computing environment. An example apparatus includes a resource manager table populator to populate a resource management table with information identifying workloads and a workload categorizer to categorize the workloads identified in the resource management table as needy workloads or offering workloads. The disclosed apparatus further include a workload selector to select an offering workload to lend resources to a needy workload, and a resource configurer. The resource configurer to cause a host included in the workload to spin a virtual machine or create a container, and assign the virtual machine or container to support the needy workload.