International Business Machines Corporation

Vikram Yadav

Pooja Malik

Gopal Kumar Bhageria

Rajesh Kumar Saxena

Sathya Santhar

Vijay Ekambaram

Harish Bharti

Sandeep Sukhija

Pranshu Tiwari

A method and associated systems use DVFS performance-scaling technology to satisfy quality-of-service performance requirements when recovering a job that had been scheduled to run on a failed virtual machine. A Buffer Time specifies a duration of time remaining, at the time of failure, for the job to complete in order to satisfy the quality-of-service requirements. Depending on relative durations of time required to repair the failed virtual machine, to perform the job on an unscaled active-mode virtual machine, and to transfer the job to another virtual machine, the system determines whether to repair the failed virtual machine or to transfer the job. If the latter, the system then determines whether to select a destination virtual machine provisioned on a DVFS-compliant platform and, if so, the system scales the DVFS-compliant platform's performance to a level sufficient to complete the job within the Buffer Time.

Reprovisioning virtual machines by means of DVFS-aware scheduling

Managing execution of a job in a computing environment. A method establishes, for a job to be executed in the computing environment, an execution plan for processing the job. The execution plan identifies computationally intensive tasks of the job and data intensive tasks of the job. The method selects a virtual machine of the computing environment to process the identified computationally intensive tasks of the job and identified data intensive tasks of the job. The method assigns the identified computationally intensive tasks of the job for foreground processing of the virtual machine and assigns the identified data intensive tasks of the job for background processing of the virtual machine. Execution of the job executes the identified computationally intensive tasks of the job in foreground processing of the virtual machine and executes the identified data intensive tasks of the job in background processing of the virtual machine.

Scheduling framework for tightly coupled jobs

Mechanisms are provided to implement an efficient translating mechanism to efficiently translating social media posts. A source language to be used to translate the social media post is identified based on words within the social media post. A highest classification is identified and the social media post is translated from the source language to a target language using a translation level associated with the highest classification. In the translation, each word and its related meaning in the target language are identified from a multi-language data structure; each word is categorized into its associated part of speech; a sentence is generated in the target language; and natural language processing is performed on each sentence in the target language to identity the existence of ambiguous connotations. Responsive to each sentence failing have any ambiguous connotations, a social medial post is generated in the target language utilizing the generated sentences.

Efficient translating of social media posts

Methods, computer program products, and systems are presented. The method computer program products, and systems can include, for instance: assigning resources of a K-tier resource pool to a certain job residing in a job queue, wherein the certain job residing in the job queue features job coupling characterized by an independent job and a dependent job which for completion depends on an output of the independent job, wherein the K-tier resource pool includes at least one foreground virtual machine (VM) having a first central processing unit (CPU) priority and at least one background virtual machine (VM) having a second CPU priority, wherein the first CPU priority is higher than the second CPU priority, wherein the assigning resources of the K-tier resource pool to the certain job includes assigning one or more foreground VM to the independent job and assigning one or more background VM to the dependent job.

K-tier architecture scheduling

A method and system for reassigning failed jobs. It is determined that a job queue of a virtual network is overloaded. Each job is set in the job queue to be processed in a scalable mode of operation as a function of the job queue being overloaded. A job is apportioned in the job queue to a virtual machine in the virtual network operating in the scalable mode of operation. The job queued by the virtual machine fails to be completed. A probability of failing to complete the job by the virtual machine is computed. It is determined, as a function of the probability of failing to complete the job, whether to complete the job queued by the virtual machine or transfer the job to a queue of a second virtual machine operating in a dynamic voltage and frequency scaling (DVFS) mode or an active mode.

Virtual machine consolidation

Method of allocating tasks in a computing environment including: receiving a software application having tasks for processing; splitting the software application into the tasks; selecting a task for processing in a first computing environment without encryption, a second computing environment with homomorphic encryption or a third computing environment without encryption based on the following algorithm: analyzing the tasks for the presence of a security marker indicating a security level of the tasks; when there is no security marker, selecting the task for processing in the least costly of first computing environment without encryption or the third computing environment without encryption; and when the security marker is present and the processing of the task involves any computation, selecting the task for processing in the least costly of the second computing environment with homomorphic encryption or the third computing environment.

Allocating tasks in a computing environment

A method and system for allocating data processing jobs between public and private cloud based on various SLA and cost factors associated to each job, and particularly, job allocation using minimal cost association by applying logistic regression. Jobs are analyzed based on various factors such as compute and operational intensity, kind of environment, I/O operations bandwidth, costs involved to deploy in private and public cloud and all these parameters are balanced to arrive at minimized cost. Methods are implemented for receiving input data representing a current request to run a job on a virtual machine, associated job characteristics, features associated with VMs running on a public networked or private networked host computing environment, and features associated with the host computing environment. The learned classifier model is run for selecting one of: the public network or private network as a host computing environment to run said current requested job on a VM resource, the learned model based on a minimized cost function.

Allocating jobs to virtual machines in a computing environment

Method of allocating tasks in a computing environment including: receiving a software application having tasks for processing; splitting the software application into the tasks; selecting a task for processing in a first computing environment without encryption, a second computing environment with homomorphic encryption or a third computing environment without encryption based on the following algorithm: analyzing the tasks for the presence of a security marker indicating a security level of the tasks; when there is no security marker, selecting the task for processing in the least costly of first computing environment without encryption or the third computing environment without encryption; and when the security marker is medium or high and the processing of the task involves any computation, selecting the task for processing in the least costly of the second computing environment with homomorphic encryption or the third computing environment.

A method and associated systems use DVFS performance-scaling technology to satisfy quality-of-service performance requirements when recovering a job that had been scheduled to run on a failed virtual machine. A Buffer Time specifies a duration of time remaining, at the time of failure, for the job to complete in order to satisfy the quality-of-service requirements. Depending on relative durations of time required to repair the failed virtual machine, to perform the job on an unsealed active-mode virtual machine, and to transfer the job to another virtual machine, the system determines whether to repair the failed virtual machine or to transfer the job. If the latter, the system then determines whether to select a destination virtual machine provisioned on a DVFS-compliant platform and, if so, the system scales the DVFS-compliant platform's performance to a level sufficient to complete the job within the Buffer Time.

Efficiently repairing virtual machines by means of DVFS-aware proactive scheduling

Systems, methods and tools for managing the job queues of virtual machines, maintaining a low energy profile and a quality of service within the contractual service agreement. The systems migrate jobs to a new VM queue when a assigned VM has failed. The systems employ machine learning techniques to make decisions whether or not to reallocate the job to a VM running in an active mode (non-scalable mode) or a VM operating under a dynamic voltage and frequency scaling (DVFS) mode. The systems reconcile job failures, transfer and/or complete jobs using the network of VMs without degrading the service quality, maintaining a lower power consumption policy through scalable modes, including idle, busy, sleep, DVFS gradient and DVFS maximum modes, improving the overall reliability of the data center by switching the jobs to scalable nodes, increasing the recoverability of the systems in the virtualized environments.

Managing handoffs between a plurality of mobile devices in a phone hosted microservices architecture in a same vicinity, with each of the mobile devices connected to a cognitive reliability engine. The cognitive reliability engine, for each of the plurality of mobile devices hosting a hosted microservice, determining a reliability score for a time period through context and activity recognition of a user owning the mobile device. Depending on the reliability score, different beacons with data packets indicating that the microservice will end, the microservice may end, or the microservice will continue with surety for a specific time period.

Cognitive reliability engine for smooth handoff in phone-hosted microservices

One or more processors determine an energy consumption need of a computing network. One or more processors lower an energy consumption level of one or more nodes in the computing network based, at least in part, on the energy consumption need. One or more processors raise the energy consumption level of at least one of the one or more nodes in the computing network in a dynamic response to a service request.

Managing data center power consumption

Provision of power scheduling framework for a plurality of virtual machines to facilitate energy savings in handing service instances includes configuring a first set of virtual machines in an active power mode and a second set of virtual machines in one or more other power modes, and managing, across a time period and based on service instance arrival, power mode configuration for the first set of virtual machines and the second set of virtual machines, the managing including initiating one or more transitions of at least one virtual machine between power modes of the plurality of power modes, wherein a transition energy cost is associated with transitioning from a lower power mode of the plurality of power modes to a higher power mode of the plurality of power modes, and wherein the managing accounts for transition energy cost in determining the one or more transitions.

Growing community of inventors

Farrukhabad, India

Vikram Yadav