Vmware, Inc.

Hewlett-packard Development Company, L.p.

Cisco Technology, Inc.

Microsoft Technology Licensing, LLC

Netapp, Inc.

Pradeep Padala

Aashish Parikh

Parth Shah

Xiaoyun Zhu

Anne Holler

Ajay Gulati

Rean Griffith

Mustafa Uysal

Lei Lu

Shankari Kalyanaraman

Arif Abdulhusein Merchant

Ganesha Shanmuganathan

Vijayaraghavan Soundararajan

Himanshu Raj

Kiran Kamity

Adarsh Jagadeeshwaran

Selvi Kadirvel

Rohit Bhoj

Pinar Yanardag Delul

Ramachandran Ramjee

Venkata N Padmanabhan

Sharad Singhal

Vishal I Gupta

Zhikui Wang

Mazda A Marvasti

Michael Kevin Larkin

Aalap Desai

Aastha Bhardwaj

Peter Varman

Badhrinath Sampathkumar

Nathaniel Morris

Mukil Kesavan

Tathagata Das

Kenneth Salem

David Lee

Neelima Mukiri

Zhefu Jiang

Peter Joseph Varman

Identifying a likely cause of a problem in an application can include: injecting a series of a set of predetermined faults into the application; sampling a subset of a set of training features from the application during each predetermined fault injected into the application and labeling each subset with an identifier of the corresponding predetermined fault; and training a classifier to identify the likely cause by associating a set of real-time features sampled from the application that pertain to the problem to one or more of the predetermined faults in response to the training features.

Diagnosing application problems by learning from fault injections

A system and method for autoscaling a multi-tier application, that has components executing on a plurality of tiers of a virtual data center, allocates resources to each of the plurality of tiers based on cost and performance. An application performance is determined, and a new application performance is estimated based at least partially on an application reservation and an application limit. An optimized utility of the application is calculated based on the cost to execute the application, the application reservation, and the application limit. A scaling factor for each tier is then determined to scale up or down a number of virtual machines operating in each of the tiers.

Automated scaling of application in virtual data centers

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.

Automatic placement of clients in a distributed computer system satisfying constraints

A system and method for performing an operational metric analysis for a virtual appliance uses application operational data from multiple instances of the virtual appliance. The application operational data is then used to generate an operational metric prediction for the virtual appliance.

Crowd-sourced operational metric analysis of virtual appliances

A placement simulator is used for testing a placement engine in a virtual machine environment. The placement simulator includes a simulation controller, an event manager, and an inventory manager. The simulation controller receives input data for a simulated datacenter. The event manager invokes event handlers for a sequence of events from the input data. The inventory manager stores states of inventory objects to simulate deployment of virtual infrastructure resources by the placement engine based on the sequence of the events.

System and method for simulating virtual machine (VM) placement in virtual datacenters

In one embodiment, a system has host machines forming a cluster. Each host machine runs containers, where each container includes a segment of hardware resources associated with the host machine, a segment of an operating system utilized by the host machine, and at least one application. Host agents operate on the host machines. Each host agent collects operational parameters associated with the containers on each host machine. A management platform is operative to divide the cluster into container pools, where each container pool includes a sub-set of computation resources in the cluster and has associated container pool metrics including a priority level and computation resource limits. Operational parameters are collected from the host agents. The operational parameters are evaluated in accordance with the container pool metrics.

Container management and application ingestion engine

An automatic scaling system and method for reducing state space in reinforced learning for automatic scaling of a multi-tier application uses a state decision tree that is updated with new states of the multi-tier application. When a new state of the multi-tier application is received, the new state is placed in an existing node of the state decision tree only if a first attribute of the new state is same as a first attribute of any state contained in the existing node and a second attribute of the new state is sufficiently similar to a second attribute of each existing state contained in the existing node based on a similarity measurement of the second attribute of each state contained in the existing node with the second attribute of the new state.

System and method for reducing state space in reinforced learning by using decision tree classification

A system and method for managing resources in a distributed computer system that includes at least one resource pool for a set of virtual machines (VMs) utilizes a set of desired individual VM-level resource settings that corresponds to target resource allocations for observed performance of an application running in the distributed computer system. The set of desired individual VM-level resource settings are determined by constructing a model for the observed application performance as a function of current VM-level resource allocations and then inverting the function to compute the target resource allocations in order to meet at least one user-defined service level objective (SLO). The set of desired individual VM-level resource settings are used to determine final RP-level resource settings for a resource pool to which the application belongs and final VM-level resource settings for the VMs running under the resource pool, which are then selectively applied.

Performance-driven resource management in a distributed computer system

Systems and methods for finding solutions exhaustively in distributed load balancing are provided. A plurality of virtual machines (VMs) is in communication with a virtual machine management server (VMMS). The VMMS is configured to generate a matrix that represents a mapping of a plurality of virtual machines (VMs) to a plurality of hosts and to calculate a first imbalance metric of the matrix. The VMMS is also configured to identify a plurality of candidate migrations the VMs. The VMMS searches through the solution space efficiently and can perform an exhaustive search to find the optimal solution. For each candidate migration, the VMMS is configured to alter the matrix to represent the candidate migration and to calculate a candidate imbalance metric based on the altered matrix. The VMMS is also configured to determine which candidate migration to perform based at least in part on the candidate imbalance metric for each candidate migration and the first imbalance metric.

Systems and methods for finding solutions in distributed load balancing

The current document is directed to an analysis subsystem within a large distributed computing system, such as a virtual data center or cloud-computing facility, that monitors the operational states associated with a multi-tiered application and provides useful information for determining one or more causes of various types of failures and undesirable operational states that may arise during operation of the multi-tiered application. In one implementation, the analysis subsystem collects metrics provided by various different types of metrics sources within the computational system and employs principal feature analysis to select a generally small subset of the collected metrics particularly relevant to monitoring a multi-tiered application and diagnosing underlying causes of operational states of the multi-tiered application. The analysis subsystem develops one or more conditional probability distributions with respect to the subset of metrics. These one or more conditional probability distributions, in turn, allow the analysis subsystem to provide useful information for analysis of the causes of failures and undesirable system states associated with the multi-tiered application.

Method and system that analyzes operational characteristics of multi-tier applications

Growing community of inventors

Sunnyvale, CA, United States of America

Pradeep Padala