Showing papers by "Marcus Fontoura published in 2018"

Predictive rightsizing for virtual machines in cloud computing systems

Abstract: Various embodiments, methods, and systems for implementing a predictive rightsizing system are provided. Predicted rightsized deployment configurations are generated for virtual machine “VM” deployments having deployment configurations that are modified to predicted rightsized deployment configurations based on a prediction engine. In operation, a VM deployment, associated with a request to deploy one or more VMs on a node, is accessed at a predictive rightsizing controller. A predicted resource utilization for the VM deployment is generated at the prediction engine and accessed at the predictive rightsizing controller. The predicted resource utilization is generated based on a prediction engine that uses past behaviors and features associated with previous VM deployments. Based on the predicted resource utilization, a predicted rightsized deployment configuration is generated for the VM deployment. The predicted rightsized deployment configuration is communicated such that the request to deploy the VMs is executed based on the predicted rightsized deployment configuration.

Availability management operations in a distributed computing system

Abstract: Various methods and systems for implementing an availability management system for implementing an availability management, in distributed computing systems, are provided. An availability management system implements an availability manager and an availability configuration interface to meet availability guarantees for tenant infrastructure. The availability management systems operates with availability zones, computing clusters, fault and upgrade domains to allocate and de-allocate virtual machine sets of virtual machine instances to a distributed computing system based on tenant-defined availability parameters. The availability parameters are used to generate an availability profile. The availability manager is configured to, based on an availability profile, allocate the virtual machine sets based an allocation scheme. The availability manager specifically performs scaling-out, scaling-in and rebalancing operations for allocating and de-allocating the virtual machine sets. An allocation configuration that meets the availability parameters of the availability profile is determined the allocation configuration is used for allocating the virtual machine set.

Availability management in a distributed computing system

Abstract: Various methods and systems for implementing an availability management system for implementing an availability management, in distributed computing systems, are provided. An availability management system implements an availability manager and an availability configuration interface to meet availability guarantees for tenant infrastructure. The availability management systems operates with availability zones, computing clusters, fault and upgrade domains to allocate and de-allocate virtual machine sets of virtual machine instances to a distributed computing system based on tenant-defined availability parameters. The availability manager is configured to: based on an availability profile, allocate the virtual machine sets across the availability zones using an allocation scheme. The allocation scheme is a virtual machine set spanning availability zones allocation scheme for performing evaluations to determine an allocation configuration defined across at least two availability zones for allocating virtual machine sets. When the allocation configuration meets the availability parameters, the allocation scheme selects the allocation configuration for allocating the virtual machine set.

Resource oversubscription based on utilization patterns in computing systems

Abstract: Techniques of managing oversubscription of network resources are disclosed herein. In one embodiment, a method includes receiving resource utilization data of a virtual machine hosted on a server in a computing system. The virtual machine is configured to perform a task. The method also includes determining whether a temporal pattern of the resource utilization data associated with the virtual machine indicates one or more cycles of resource utilization as a function of time and in response to determining that the temporal pattern associated with the virtual machine indicates one or more cycles of resource utilization as a function of time, causing the virtual machine to migrate to another server that is not oversubscribed by virtual machines in the computing system.

Update log assigned by cloud service instance

Abstract: A computing device assigns instances to the cloud network. An allocation application stored in a memory and executed by a processor is configured to receive a request to allocate an instance to a nodein the cloud network. The instances are defined by a set of characteristics. The allocation application is configured to create a request equivalence class based on the set of characteristics. One ormore node equivalence classes for the request equivalence class are cached at a high speed. A node update log comprising a list of a plurality of modified nodes is maintained. A request equivalent class is selected in response to a request for assigning a new instance comprising the set of characteristics. Communication with the node update log is performed to identify the modified node in the one or more node equivalence classes and update the modified node before the assigned node is selected in the one or more node equivalence classes.

Availability management interfaces in a distributed computing system

Abstract: Various methods and systems for implementing an availability management system for implementing an availability management, in distributed computing systems, are provided. An availability management system implements an availability manager and an availability configuration interface to meet availability guarantees for tenant infrastructure. The availability management systems operates with availability zones, computing clusters, fault and upgrade domains to allocate and de-allocate virtual machine sets of virtual machine instances to a distributed computing system based on tenant-defined availability parameters. The availability configuration interface of the availability management system supports receiving availability parameters that are used to generate an availability profile. The availability profile comprises availability parameters including, configurations for spanning or non-spanning, rebalancing, fault domains, update domains, and availability zones associated with allocating a virtual machine sets. The availability zones are exposed to the tenant as logically-defined availability zones that map to physically-defined availability zones for an abstracted implementation of the availability zones.

Effective cloud service instance using caching and sampling

Abstract: An effective cloud service instance using caching and sampling includes the following steps that: a computing device assigns an instance to nodes in a cloud network; an allocation application that isstored in a memory and executed by a processor is configured to receive a request and further allocate the instance defined by a set of characteristics; if a request equivalence class exists, the setof characteristics of the instance are compared to each set of characteristics corresponding to an R request equivalence class, wherein each R request equivalence class includes one or more node equivalence classes, and each node equivalence class includes a plurality of nodes; if the set of characteristics of the instance fail to match, the allocation application is configured to create a new request equivalence class; and if the set of characteristics of the instance match successfully, the allocation application is configured to identify a selected node corresponding a matched R request equivalence class in the R request equivalence classes from the one or more node equivalence classes.