At a first compute instance run on a virtualization host, a local instance scaling manager is launched. The scaling manager determines, based on metrics collected at the host, that a triggering condition for redistributing one or more types of resources of the first compute instance has been met. The scaling manager causes virtualization management components to allocate a subset of the first compute instance's resources to a second compute instance at the host.

Aspects of the disclosure relate to a real-time dynamic container optimization computing platform. The real-time dynamic container optimization computing platform may receive a request to create a first processing block and first data associated with the first processing block. The real-time dynamic container optimization computing platform may utilize a plurality of models to select a first computing device for the first processing block. The real-time dynamic container optimization computing platform may generate and deploy a container to the first computing device. The real-time dynamic container optimization computing platform may monitor execution of the container on the first computing device. The real-time dynamic container optimization computing platform may migrate the container to the second computing device if an issue with execution of the container on the first computing device is detected.

Techniques are described for sharing prepopulated container image caches among container execution environments to improve the performance of container launches. The container images used to prepopulate such a cache at a computing device supporting one or more container execution environments can include various container images that are used as the basis for a wide range of user-created containers such as, for example, container images representing popular operating system distributions, database servers, web-application frameworks, and so forth. Existing systems typically obtain these container images as needed at runtime when launching containers (for example, from a container registry or other external source), often incurring significant overhead in the container launch process. The use of a prepopulated container image cache can significantly improve the performance of container launches by making such commonly used container images available to container execution environments running at a computing device ahead of time.

Deferred reclaiming of secure guest resources within a computing environment is provided, which includes initiating, by a host of the computing environment, removal of a secure guest from the computing environment, while leaving one or more resources of the secure guest to be reclaimed asynchronous to the removal of the secure guest. The deferring also includes reclaiming the one or more secure guest resources asynchronous to the removal of the secure guest, where the one or more secure guest resources are available for reuse as the one or more secure guest resources are reclaimed asynchronous to the removal of the secure guest.

Upon receiving a request to hibernate a hypervisor of a virtualization system running on a first computer, acts are carried out to capture a state of the hypervisor, where the state of the hypervisor comprises hypervisor logical resource parameters and an execution state of the hypervisor. After hibernating the hypervisor by quiescing the hypervisor and storing the state of the hypervisor into a data structure, the data structure is moved to a different location. At a later moment in time, the data structure is loaded onto a second computing machine and restored. The restore operation restores the hypervisor and all of its state, including all of the virtual machines of the hypervisor as well as all of the virtual disks and other virtual devices of the virtual machines. Differences between the first computing machine and the second computing machine are reconciled before execution of the hypervisor on the second machine.

A cloud network is a complex environment in which hundreds and thousands of users or entities can each host, create, modify, and develop multiple virtual machines. Each virtual machine can have complex behavior unknown to the provider or maintainer of the cloud. Technologies disclosed include methods, systems, and apparatuses to monitor the complex environment to detect network anomalies using machine learning techniques. In addition, techniques to modify and adapt to user feedback are provided allowing the developed models to be tuned for specific use cases, virtual machine types, and users.

An information handling system may include at least one processor; and a non-transitory memory coupled to the at least one processor. The information handling system may be configured to: execute a plurality of virtual machines having workloads associated therewith; during selected times, apply a plurality of configuration settings relating to the at least one processor while executing the workloads of the plurality of virtual machines; track a plurality of performance metrics relating to the at least one processor during the selected times; and predictively determine a selected one of the plurality of configuration settings that is predicted to improve performance of the workloads.

A method of restarting a virtual machine (VM) running in a cluster in a first data center, in a second data center, includes: transmitting images of VMs, including a first VM, running in the cluster of hosts at a first point in time to the second data center for replication in the second data center; generating difference data representing a difference in an image of the first VM at a second point in time and the image of the first VM at the first point in time; transmitting the difference data to the second data center; setting the first VM to be inactive in the first data center; and communicating with a control plane in the second data center to set as active, and power on, a VM in the second data center using the replicated image of the first VM updated with the difference data.

A cloud management server and method for performing automatic placement of clients in a distributed computer system uses a list of compatible clusters to select an affinity cluster to place the clients associated with an affinity constraint. As part of the placement method, a cluster that cannot satisfy any anti-affinity constraint associated with the clients and the affinity constrain is removed from the list of compatible clusters. After the affinity cluster has been selected, at least one cluster in the distributed computer system is also selected to place clients associated with an anti-affinity constraint.

A multi-phase boot operation of a virtualization manager at a virtualization host is initiated at an offload card. In a first phase of the boot, a security key stored in a tamper-resistant location of the offload card is used. In a second phase, firmware programs are measured using a security module, and a first version of a virtualization coordinator is instantiated at the offload card. The first version of the virtualization coordinator obtains a different version of the virtualization coordinator and launches the different version at the offload card. Other components of the virtualization manager (such as various hypervisor components that do not run at the offload card) are launched by the different version of the virtualization controller.