International Business Machines Corporation

Airbnb, Inc.

Harigovind Venkatraj Ramasamy

Ruchi Mahindru

John Alan Bivens

Valentina Salapura

Min Li

Eugen Schenfeld

Yaoping Ruan

Soumitra Sarkar

Long Wang

Yu Deng

Koushik Kumar Das

Rafah A Hosn

Mahesh Viswanathan

Norbert George Vogl

Richard Edwin Harper

Michael Man Behrendt

Kaoutar El Maghraoui

James William Murray

Nancy L Burchfield

Russell W Bergs

Matthew R Koozer

Rongda Zhu

Nathaniel Hang

Maja Vukovic

Jinho Hwang

Birgit Monika Pfitzmann

Salman Abdul Baset

Chunqiang Tang

Rohit Ranchal

Lakshminarayanan Renganarayana

Jill L Jermyn

Cuong Manh Pham

Uttam Thakore

Su Liu

Ashish Kundu

Yu Gu

Karthick Rajamani

Richard Harold Boivie

Fan Jing Meng

Elliot Karl Kolodner

John Bruce Carter

Howard Neil Anglin

Nikolai A Joukov

Robert Filepp

Tian Wu

Brian L Peterson

Yi-hsiu Wei

Kun Bai

Eric James Barkie

Kamal Bhattacharya

Byung Chul Tak

Jing Luo

Pei Ni Liu

Ravithej Chikkala

Olivia Choudhury

Simon J Kofkin-Hansen

Colm V Malone

Cheng-Lin Tsao

Kewei Sun

Zi Xiao Zhu

Kevin David Galloway

Karen Cheng

Amos A Omokpo

Sandhya Kapoor

Sandhya Narayan

Manas Singh

Amarendra Das

Saurabh Jha

Theodor Razvan Ionescu

Christopher Moss

Richard Christopher Smith

Abhas Bodas

James R S Barros, Ii

David L Westerman

Christopher C Young

Embodiments of the present invention relate to methods, systems, and computer program products for event management. In a method, a plurality of notes that are comprised in a plurality of event records are obtained in a computer system. A plurality of paragraphs that are comprised in the plurality of notes are classified into a plurality of content types based on a content analysis of the plurality of paragraphs. The plurality of notes are classified into a plurality of semantic types based on the plurality of content types and a syntactic parsing to the plurality of notes. A knowledge item is generated for managing an event in the computer system based on a group of notes in the plurality of notes that are classified into the plurality of semantic types. With these embodiments, knowledge items for managing events may be obtained in an easier and more effective way.

Event management in computer system

A method, apparatus, system, and computer program product for training a global machine learning model. A hierarchical structure for nodes in which the global machine learning model is located at a primary node in the hierarchical structure is identified. Authorized nodes in which local data is authorized for use in training in the authorized nodes for a local training of local machine learning models are determined. The machine learning models in the authorized nodes are trained using the local data in the authorized nodes to generate local model updates to weights in the local machine learning models. The local model updates to the weights are propagated upward in the hierarchical structure to the global machine learning model, wherein a node receiving local model updates to the weights from nodes from a lower level aggregates the weights in the local model updates received from the nodes in the lower level.

Hierarchical federated learning using access permissions

Various embodiments for optimizing memory bandwidth in a disaggregated computing system, by a processor device, are provided. Respective memory devices are assigned to respective processor devices in the disaggregated computing system, the disaggregated computing system having at least a pool of the memory devices and a pool of the processor devices. An analytic function is performed on data resident in the pool of the memory devices using memory bandwidth not currently committed to a primary compute task.

Opportunistic data analytics using memory bandwidth in disaggregated computing systems

A base query having a plurality of base query terms is obtained. A plurality of problem log files are accessed. Words, contained in a corpus vocabulary, are extracted from the plurality of problem log files. Based on the words extracted from the plurality of problem log files, a first expanded query is generated from the base query. The corpus is queried, via a query engine and a corpus index, with a second expanded query related to the first expanded query.

Log-aided automatic query expansion approach based on topic modeling

A processor may receive information about one or more environmental factors. The processor may predict, based on the one or more environmental factors, that a particular datacenter in a distributed computing environment will experience elevated bit-flip error rates during a certain time period. The processor may select, based on the predicted elevated bit-flip error rates, one or more specific applications in the particular datacenter to be protected. The processor may protect the selected one or more specific applications during the certain time period of the predicted elevated bit-flip rates.

Application protection from bit-flip effects

Respective memory devices are assigned to respective processor devices in a disaggregated computing system, the disaggregated computing system having at least a pool of the memory devices and a pool of the processor devices. An iterative learning algorithm is used to define data boundaries of a dataset for performing an analytic function on the dataset simultaneous to a primary compute task, unrelated to the analytic function, being performed on the dataset in the pool of memory devices using memory bandwidth not currently committed to the primary compute task, thereby efficiently employing the unused memory bandwidth to prevent underutilization of the pool of memory devices.

Server resources in a data center are disaggregated into shared server resource pools, including an accelerator (e.g., FPGA) pool. Servers are constructed dynamically, on-demand and based on workload requirements, by allocating from these resource pools. According to this disclosure, accelerator utilization in the data center is managed proactively by assigning accelerators to workloads in a fine granularity and agile way, and de-provisioning them when no longer needed. In this manner, the approach is especially advantageous to automatically provision accelerators for data analytic workloads. The approach thus provides for a 'micro-service' enabling data analytic workloads to automatically and transparently use FPGA resources without providing (e.g., to the data center customer) the underlying provisioning details. Preferably, the approach dynamically determines the number and the type of FPGAs to use, and then during runtime auto-scales the FPGAs based on workload.

Utilizing accelerators to accelerate data analytic workloads in disaggregated systems

An embodiment includes extracting statistical data associated with invocation of an application programming interface (API) from a log and using the statistical data to calculate a performance value and generate an aggregate dataset that combines the performance value with performance values associated with other invocations of the API. The embodiment includes calculating metric values for performance values for respective time intervals of a time period and calculating mean and standard deviation values of the metric values for the time period. The embodiment includes selecting the API as a candidate API and detecting a Customer Impacting Event (CIE) by applying a machine learning algorithm using monitored values associated with the candidate API during a time frame defined by a rolling window. The embodiment also includes automatically initiating a selected alert from among a plurality of alert options based at least in part on the monitored values associated with the CIE.

Automated selection of performance monitors

A method, computer program product, and computer system for prioritization of data collection and analysis for incident detection. Data points within a plurality of monitor data corresponding to effects of incidents may be identified by a computing device. The plurality of monitor data may be converted to a set of time series extracted from the plurality of monitor data. The set of time series may be aligned and summarized. A causal relationship graph may be constructed between the set of time series extracted from the plurality of monitor data. Features may be prioritized by likelihood of importance in incident analysis based upon, at least in part, the causal relationship graph.

Prioritization of data collection and analysis for incident detection

A new approach to resiliency management is provided in a data center wherein servers are constructed dynamically, on-demand and based on workload requirements and a tenant's resiliency requirements by allocating resources from these pools. In this approach, a set of functionally-equivalent 'interchangeable compute units' (ICUs) are composed of resources from resource pools that have been extended to include not only different resource types (CPU, memory, accelerators), but also resources of different specifications (specs) and flavors. As a workload is being processed, the health or status of the resources are monitored. Upon a performance issue or failure event, a resiliency manager can swap out a current ICU and replace it with a functionally-equivalent ICU. Preferably, individual ICUs are hosted on one of: resources of a same type each with different specifications, and resources of a same type and specification and different flavors. The approach enables failure independence in a disaggregated environment.

Building a highly-resilient system with failure independence in a disaggregated compute environment

A method, apparatus, system, and computer program product for facilitating evidence collection. A set of evidence requirements is identified by computer system in which the set of evidence requirements is for a control that manages a set of resources in the computer system to enforce a policy in the computer system. Labels are associated by the computer system with historical evidence for the set of requirements. The historical evidence comprises prior evidence collected for compliance with the set of evidence requirements for the control and was accepted to meet the set of evidence requirements for the control. The historical evidence with the labels form labeled historical evidence. Rules for mining evidence for the set of evidence requirements for the control using the labeled historical evidence are learned by a machine learning model in the computer system.

Evidence mining for compliance management

A method of organizing computer resources includes receiving a specification defining a plurality of quiescence groups of independent component instances for each of at least two services, and performing a first load balancing of the quiescence groups across a plurality of physical servers to define a plurality of supergroups while assigning each of the physical servers across the supergroups.

Hardware placement and maintenance scheduling in high availability systems

For power management in a disaggregated computing system, upon detecting an electrical power reduction to the disaggregated computing system, respective workloads being performed by respective processors within a set of processors are prioritized according to a service level agreement (SLA) of the respective workloads. Backup electrical power is dynamically allocated to the respective processors performing the respective workloads based upon a priority of the SLA of the respective workloads.

SLA-based backup power management during utility power interruption in disaggregated datacenters

Various embodiments for performing hardware upgrades in a disaggregated computing environment. A workload is run on a disaggregated computing system while providing a new component to at least one of a plurality of component pools used by the disaggregated computing system. Point-to-point circuit wire level switching is used to switch the disaggregated system from an assigned component residing in a first of the plurality of component pools to the new component residing in a second of the plurality of component pools without interrupting the running workload.

Live, in-line hardware component upgrades in disaggregated systems

Server resources in a data center are disaggregated into shared server resource pools. Servers are constructed dynamically, on-demand and based on workload requirements, by allocating from these resource pools. A disaggregated compute system of this type keeps track of resources that are available in the shared server resource pools, and it manages those resources based on that information. Each server entity built is assigned with a unique server ID, and each resource that comprises a component thereof is tagged with the identifier. As a workload is processed by the server entity, its composition may change, e.g. by allocating more resources to the server entity, or by de-allocating resources from the server entity. Workload requests are associated with the unique server ID for the server entity. When a workload request is received at a resource, it matches its unique server ID to that of the request before servicing the request.

Specifying a disaggregated compute system

Growing community of inventors

Ossining, NY, United States of America

Harigovind Venkatraj Ramasamy