Emc IP Holding Company LLC

Dell Products Inc.

Gopal Singh

Rahul Deo Vishwakarma

Shelesh Chopra

Parmeshwr Prasad

Jayanth Kumar Reddy Perneti

Mahantesh M Ambaljeri

Hemant Ramesh Gaikwad

Sujan Kumar Shetty

Pooja Singh

A method and system for health rank based virtual machine restoration using a conformal framework. With respect to enterprise data protection, solutions need to address two primary responsibilities: at the onset of failure or disaster, restore any running applications, as well as any lost or damaged data; and minimize (if not eliminate) the future occurrence of such failures or disasters. In tackling the latter responsibility, the disclosed method and system leverage health-score assessments to ensure the restoration of virtual machines onto healthy infrastructure. The aforementioned health-score assessments employ clustering to identify, and a conformal framework to rank, healthy hosts onto which virtual machines may be restored.

Method and system for health rank based virtual machine restoration using a conformal framework

Methods, apparatus, and processor-readable storage media for analyzing time series data for sets of devices using machine learning techniques are provided herein. An example computer-implemented method includes processing time series data from multiple devices; generating at least one data forecast by applying, in response to a request from at least one user, one or more machine learning techniques to at least a portion of the processed time series data; computing one or more qualifying values attributable to the at least one generated data forecast by providing the at least one generated data forecast and the at least a portion of the processed time series data to a conformal prediction framework; and performing one or more automated actions based at least in part on the at least one generated data forecast and the one or more computed qualifying values.

Analyzing time series data for sets of devices using machine learning techniques

Embodiments are described for an autonomously and dynamically allocating resources in a distributed network based on forecasted a-priori CPU resource utilization, rather than a manual throttle setting. A multivariate (CPU idle %, disk I/O, network and memory) rather than single variable approach for Probabilistic Weighted Fuzzy Time Series (PWFTS) is used for forecasting compute resources. The dynamic throttling is combined with an adaptive compute change rate detection and correction. A single spike detection and removal mechanism is used to prevent the application of too many frequent throttling changes. Such a method can be implemented for several use cases including, but not limited to: cloud data migration, replication to a storage server, system upgrades, bandwidth throttling in storage networks, and garbage collection.

System and method for autonomous and dynamic resource allocation in storage systems

Methods, apparatus, and processor-readable storage media for implementing dual relationship-based hash structures for non-volatile memory technology are provided herein. An example computer-implemented method includes generating a hash storage structure by: creating multiple layers of storage components, wherein a first layers comprise a first number of storage components and at least a second layer comprises at least a second number of storage components less than the first number of storage components; configuring the at least a second layer to share at least a portion of the storage components therein with at least a portion of the storage components in the first layer; partitioning each of the storage components in each of the multiple layers into multiple portions; configuring a respective status value space for each of the multiple portions of each of the storage components in the multiple layers; and implementing the hash storage structure in at least one storage system.

Dual relationship-based hash structure for non-volatile memory technology

Methods, apparatus, and processor-readable storage media for automatically allocating device resources using machine learning techniques are provided herein. An example computer-implemented method includes determining values for multiple parameters from data obtained from one or more devices; generating at least one device component failure prediction by applying one or more machine learning techniques to at least a portion of the determined values; computing one or more qualifying values attributable to the at least one generated device component failure prediction by providing the at least one generated device component failure prediction to a conformal prediction framework; and automatically allocating one or more resources to at least a portion of the one or more devices based at least in part on the at least one generated device component failure prediction and the one or more computed qualifying values.

Automatically allocating device resources using machine learning techniques

Drives of a RAID group are classified as either healthy or failing using a trained learning process. The failure probability is then determined for each failing drive using a Venn-ABERS framework which provides a boundary range on the failure prediction probability. A variable sparing mechanism is used to enable one or more drives of the RAID group to be used as dual-purpose drives. In a first state, the dual-purpose drives are user-addressable drives and are available to be used to process IO workload on the RAID group. Spreading the IO workload on the RAID group across a larger number of drives results in increased performance in the form of reduced latency. In a second state, the dual-purpose drives are not user-addressable and are spare drives in the RAID group, which improves the level of protection provided to data stored in the RAID group.

Growing community of inventors

Lucknow, India

Gopal Singh