International Business Machines Corporation

Kyndryl, Inc.

Jean Xu Yu

James Patrick Smith

Avishai Haim Hochberg

Hang Xiao

Ilyas Mohamed Iyoob

Emrah Zarifoglu

Omar Odibat

Nan Fan

Allan Bradley Winslow

Ting Bin Wu

Ping Wang

David Maxwell Cannon

Colin Scott Dawson

Kevin Lee Gibble

Robert Clair Edwards, Jr

Angelo Danducci, Ii

Martine Bruce Wedlake

David George Derk

Ming Yan

James J Myers

Joanne T Nguyen

Neil Gregory Rasmussen

Kenneth Eugene Hannigan

Thomas Edward O'Brien

Nduwuisi Emuchay

Yan Xiao

Tai Woo Nam

Peter B Symonds

Gergana Vassileva Markova

Keri Wheatley

Chuan Liu

Thu Nguyen

Sean P Cox

Jawed Iqbal

Steven V Voyk

An embodiment includes generating an input document that includes a plurality of text fields of attribute data. The embodiment also includes extracting a set of candidate features from the attribute data using a feature extraction module that evaluates the attribute data using first and second machine learning models, where the first machine learning model scores terms in the input document and the second machine learning model includes a deep learning model. The embodiment also includes calculating feature-selection values for respective features of the set of candidate features and generating tag information for a remote computing asset using a machine learning classifier that predicts the tag information based on the feature-selection values.

Methods and apparatus for feature generation using improved term frequency-inverse document frequency (TF-IDF) with deep learning for accurate cloud asset tagging

Provided are techniques for exploiting object tags in order to produce a work order across several backup engines for a backup job. A selection of tags for an object is received. A work order for the object is created using the tags. The work order is stored with a virtual server that stores the object, wherein a data mover of a backup engine retrieves the work order from the virtual server to back up the object.

Exploiting object tags to produce a work order across backup engines for a backup job

Feature generation for asset classification

An embodiment includes calculating an efficacy value of a context assigned to an asset, the efficacy value being based at least in part on a percentage of assets assigned to the context. The embodiment also includes calculating, responsive to the calculated efficacy value being within a predetermined range, a normalized relevance value of a first tag associated with the asset. The embodiment also includes generating relative relevance value for the first tag and a reason code associated with a basis for the relative relevance value of the first tag based at least in part on the normalized relevance value of the first tag. The embodiment also includes initiating, responsive to generating the relative relevance value and the reason code for the first tag, a database command on a training dataset stored in the database that includes the first tag.

Ranking of asset tags

During a data protection operation, a system exploits a virtual hierarchy to centralize the configuration and management of operating system credentials of numerous virtual guests. For each virtual guest, the system uses the credential to collect a single Globally Unique Identifier (GUID) previously generated and stored in-guest by any data protection agent. The system stores the collected GUID as a custom property in the context of the virtual hierarchy. The system also exploits the virtual hierarchy custom properties to determine if GUIDs are copies due to virtual guest replication. The system ensures GUID uniqueness by requesting regeneration of the GUID by in-guest data protection agents. Using GUIDs that are unique across the virtual hierarchy, the system can correlate application data of multiple in-guest data protection agents.

Data protection backup agent management

Methods, systems, and products are disclosed for correlating business workflows with transaction tracking, the method including identifying an instrumentation point in a business process of a business workflow that invokes an IT resource that carries out at least a portion of the business process and associating, at the instrumentation point, the business process and the IT resource with a transaction name having business semantics such that correlation data gathered about the IT resource may be linked with the business process. Correlating business workflows with transaction tracking may also include gathering correlation data about the IT resource, selecting one of the plurality of service specific correlator passing methods, and passing a correlator in accordance with the selected service specific correlator passing methods.

Correlating business workflows with transaction tracking

A mechanism is provided in a data processing system for de-duplication in a Web services environment. Responsive to receiving a signature for a portion of a request message from a client, the mechanism performs a query in a bit stream repository to determine whether the portion of the request message has been previously received. Responsive to determining the portion of the message has been previously received, the mechanism retrieves the portion of the message from the bit stream repository. The mechanism notifies the client that the portion of the message has been previously received such that the client does not send the portion of the message.

Data de-duplication in service oriented architecture and web services environment

During execution of an existing scheduling computer program on a client node, an update computer program and a self-describing automatic installation package are downloaded to the client node from a logical depot node implemented on an existing management server. Therefore, advantageously, no physical depot node or other additional computing device is needed for the client node to update itself. Execution of the update computer program is spawned on the client node from the existing scheduling computer program. As such, the update computer program inherits root access to the client node and security credentials to the management server from the scheduling computer program&#x2014;advantageously, then, a user does not have to perform any laborious configuration of the client node in order to update the node. The client node ultimately updates itself using the self-describing automatic installation package, which includes all the information needed for the client node to update itself.

Updating client node of computing system

A backup program running on a computing device receives a search query. The backup program transmits the search query to a search program. The backup program receives search results corresponding to the search query from the search program. The search results include a list of data objects stored on a storage device and that satisfy the search query. The backup program backs up one or more selected data objects of the list of data objects. The backup program can periodically and automatically without user interaction retransmit the search query to the search program, and receive updated search results corresponding to the search query from the search program. The updated search results include an updated list of data objects stored on the storage device that satisfy the search query. The backup program can then back up the data objects within the updated list of data objects.

Backing up data objects identified by search program and corresponding to search query

Using microblogging to dynamically track event flow of a composite enterprise application by reporting an enterprise application event. A client server detects an event of the complex enterprise application, and encodes the event data in a microblog compatible format. The client computer creates an event post message containing the event data and uploads the event post to a microblog server via the Internet. The microblog server then displays the post message in a user readable format.

Microblogging based dynamic transaction tracking for composite application flow

A computer implemented method, data processing system, and computer program product for Java class automatic deployment using byte code instrumentation technology. One or more classloaders are instrumented with byte code instrumentation code such that a class loading event is received when a class is loaded. If a determination is made that new byte code instrumentation code needs to be loaded with the loaded class, candidate classloaders that load import classes of the new byte code instrumentation code are determined. A correct classloader from the candidate classloaders to load the new byte code instrumentation code is calculated. The correct classloader is instrumented to have an extended classpath, wherein the new byte code instrumentation code is inserted into the extended classpath of the correct classloader. The class is loaded from the extended classpath and original classpath of the correct classloader.

Automatic deployment of Java classes using byte code instrumentation

A computer implemented method, data processing system, and computer program product for automatic discovery of the Java classloader delegation hierarchy. One or more classloaders are instrumented with byte code instrumentation code such that a delegation event is received each time a classloader delegates a resource finding method to a parent classloader of the classloader. From each delegation event received, the classloader that generated the delegation event and a parent classloader to which the classloader directly delegates the resource finding method is determined. Based on the delegation events received, the classloaders to which a current classloader in the one or more classloaders can directly delegate are determined. The order of delegation to the classloaders may then be determined.

Automatic discovery of the java classloader delegation hierarchy

A method for minimizing the message flow wait time for management user exits in a message broker application. A message broker application processes a request in a request and a response message flow. The request message flow generates a request identifier, collects information about the request message flow, and stores the request identifier and information in a global data map. The response message flow uses the request identifier to access the map and read the collected information without having to acquire a lock on the map. The response message flow also collects information about the response message flow, and generates management information about the request based on the information about the request message flow and the information about the response message flow. A dedicated clean up thread in the message broker application is used to remove used items from the global data map.

Minimizing message flow wait time for management user exits in a message broker application

Computer implemented method, system and computer usable program code for end-to-end transaction tracking of long-running processes such as a Business Process Execution Language (BPEL) process in a data processing system, such as a data processing system implemented in a Service-Oriented Architecture (SOA). A computer implemented method for tracking and correlating transactions in a process in a data processing system includes gathering instance identification data and instance state data regarding a transaction and storing the instance identification data and instance state data in association with transaction data regarding the transaction. Asynchronous event data is collected, and transaction performance information for an asynchronous event within the transaction is calculated using the transaction data. A tracking report of the transaction is then generated using the calculated transaction performance information.

End-to-end tracking of asynchronous long-running business process execution language processes

A mechanism is provided for automatic response time measurement of Lightweight Directory Access Protocol (LDAP) server operations are provided. The mechanism makes use of an LDAP server plug-in interface and LDAP server plug-in modules to automatically inject response time measurement code before and after each LDAP server operation during real time operation of the LDAP server in a runtime environment. In addition, the mechanism makes use of correlation tokens in the LDAP environment so that LDAP client and server operations may be correlated with one another in an end-to-end transaction. Mechanisms for passing correlation tokens between plug-in modules for the same LDAP server operation and passing correlation tokens between the LDAP client and LDAP server, and among LDAP client calls within a single session, are also provided.

Growing community of inventors

Austin, TX, United States of America

Jean Xu Yu