Abstract: Systems and methods for estimating a random distribution for an overall metric for a composite node, the composite node comprising a plurality of nodes. For each data atom of a plurality of data atoms being input to the composite node, and for each node of the plurality of nodes, at least one value may be generated for a per-node metric with respect to the data atom. A value for the overall metric with respect to the data atom may be generated based on the per-node metric values of the plurality of nodes. At least one parameter of the random distribution for the overall metric for the composite node may be estimated based on the overall metric values with respect to the plurality of data atoms.

Abstract: The present disclosure relates to systems and methods for an intelligent assistant (e.g., a chatbot) that can be used to enable a user to generate a machine learning system. Techniques can be used to automatically generate a machine learning system to assist a user. In some cases, the user may not be a software developer and may have little or no experience in either machine learning techniques or software programming. In some embodiments, a user can interact with an intelligent assistant. The interaction can be aural, textual, or through a graphical user interface. The chatbot can translate natural language inputs into a structural representation of a machine learning solution using an ontology. In this way, a user can work with artificial intelligence without being a data scientist to develop, train, refine, and compile machine learning models as stand-alone executable code.

Abstract: An online system performs predictions for real-time tasks and near real-time tasks based on available network bandwidth. A client device receives a regression based machine learning model. Responsive to receiving a task, the client device determines an available network bandwidth for the client device. If the available network bandwidth is below a threshold, the client device uses the regression based machine learning model to perform the task. If the client device determines that the network bandwidth is above the threshold, the client device extracts features of the task, serializes the extracted features, and transmits the serialized features to an online system, causing the online system to use a different machine learning model to perform the task based on the serialized features.

Abstract: A quality evaluation apparatus includes: a recording unit configured to record information indicating a state of a web browser in a storage unit at a plurality of timings; and a count unit configured to count the number of times reload operation is performed as an evaluation value of quality experienced by a user with reference to the storage unit, so that validity of the evaluation of experienced quality of web browsing is improved.

Abstract: Described are systems, methods, and techniques for collecting, analyzing, processing, and storing time series data and for evaluating and dynamically estimating a resolution of one or more streams of data points and updating an output resolution. Responsive to receiving a stream of data points, a data resolution can be derived and an output resolution can be set to a first value. When a change to the data resolution is detected, the output resolution can be changed, modifying a frequency at which output data points are generated and/or transmitted. In some instances, a detector can be implemented to trigger an alert responsive to ingested data points corresponding with triggering parameters. An output resolution for the detector can be dynamically modified based on dynamically detecting a change to the data resolution of the stream of data.

Abstract: An improved decentralized, blockchain-driven network for artificial intelligence (AI)-enabled skills exchange between Intelligent Personal Assistants (IPAs) in a network is disclosed that is configured to perform computational tasks or services (also referred to herein as “skills”) in an optimally-efficient fashion. In some embodiments, this may comprise a first IPA paying an agreed cost to a second IPA to perform a particular skill in a more optimally-efficient fashion. In some embodiments, a skills registry is published, comprising benchmark analyses and costs for the skills offered by the various nodes on the skills exchange network. In other embodiments, a transaction ledger is maintained that provides a record of all transactions performed across the network in a tamper-proof and auditable fashion, e.g., via the use of blockchain technology. Over time, the AI-enabled nodes in the system may learn to scale, replicate, and transact with each other in an optimized—and fully autonomous—fashion.

Abstract: Methods, apparatus, systems and articles of manufacture for providing query selection systems are disclosed. An example query selection system includes a processor to: analyze a graph database; identify respective ones of objects associated with the graph database; obtain properties associated with the objects; identify common properties present in the respective ones of the objects; in response to determining the common properties present in the identified objects, output a list of the common properties corresponding to the respective ones of the objects; generate a table for the common properties and the respective ones of the objects; and populate the table with the common properties and the respective ones of the objects from the graph database to establish a relational database. The system further includes a machine learning model classifier to enable relational database query searching via the relational database.

Abstract: A performance monitoring system includes a metric collector configured to receive, via metric exporters, telemetry data comprising metrics related to a network of computing devices. A metric time series database stores related metrics. An alert rule evaluator service is configured to evaluate rules using stored metrics. The performance monitoring system may include a machine learning module and is configured to determine optimized metric collection sampling intervals and rule evaluation intervals, and to automatically determine recommended alert rules.

Abstract: Methods, apparatus, and processor-readable storage media for automated storage system performance validation are provided herein. An example computer-implemented method includes obtaining user input pertaining to performance testing associated with one or more storage systems deployed in at least one user environment; generating, based at least in part on the obtained user input, a storage system performance test deployment environment comprising at least one controller and at least one data collection mechanism; executing, using the storage system performance test deployment environment, at least one performance test on at least a portion of the one or more storage systems deployed in the at least one user environment; and performing one or more automated actions based at least in part on results from the at least one executed performance test.

Abstract: A computing device includes a processor and a machine-readable storage storing instructions. The instructions are executable by the processor to: determine a completed amount of data transferred for each of a plurality of data transfer jobs, each of the plurality of data transfer jobs to transfer data to a storage system; determine an estimated probability of failure for each of the plurality of data transfer jobs; and allocate computing resources of the storage system to the plurality of data transfer jobs based on the completed amount of data transferred and the estimated probability of failure of each of the plurality of data transfer jobs.

Abstract: According to one general aspect, a system may include a plurality of data nodes. Each data node may include either or both of a first storage medium and a second storage medium that is slower than the first storage medium. Each data node may be configured to store a piece data in either the first storage medium or the second storage medium. The system may be configured to store a plurality of copies of an active piece of data within two or more data nodes. A fast copy of the active piece of data may be stored by a first storage medium of a first data node. One or more slow copies of the active piece of data may be stored by respective second storage mediums of one or more respective other data nodes.

Abstract: An example system includes a memory, a processor in communication with the memory, and a supervisor. The supervisor is configured to allocate a memory space in the memory to a workload executing on the processor. The supervisor is configured to store data written by the workload as dirty memory in the memory space at least until the data is written back to a data storage. Based on a type of the workload being a first type, the supervisor is configured to trigger write back of at least a portion of the dirty memory into the data storage in response to the dirty memory exceeding a threshold level. Based on the type of the workload being a second type, the supervisor is configured to delay write back of the dirty memory into the data storage in response to the dirty memory exceeding the threshold level.

Abstract: In a computing environment, a set of executing processes each having associated resources are provided. Aggregate resources for the computing environment include multiple different types of resources. A utilization level for each of the resources within the computing environment is evaluated to determine an unconsumed capacity for each of the resources below a utilization threshold. The utilization threshold is resource-dependent. An indication of at least a portion of unconsumed capacity for each of the resources below the utilization threshold is gathered. The unconsumed portion for each of the resources below the utilization threshold is exposed for consumption by other executing processes.

Abstract: A technique for operating a cache is disclosed. The technique includes recording access data for a first set of memory accesses of a first frame; identifying parameters for a second set of memory accesses of a second frame subsequent to the first frame, based on the access data; and applying the parameters to the second set of memory accesses.

Abstract: The present disclosure relates to systems and methods for an intelligent assistant (e.g., a chatbot) that can be used to enable a user to generate a machine learning system. Techniques can be used to automatically generate a machine learning system to assist a user. In some cases, the user may not be a software developer and may have little or no experience in either machine learning techniques or software programming. In some embodiments, a user can interact with an intelligent assistant. The interaction can be aural, textual, or through a graphical user interface. The chatbot can translate natural language inputs into a structural representation of a machine learning solution using an ontology. In this way, a user can work with artificial intelligence without being a data scientist to develop, train, refine, and compile machine learning models as stand-alone executable code.

Abstract: A tracing controller may utilize a binary execution trace mechanism to trace execution of compiled application machine code. The tracing controller may initiate hardware tracing to gather control-flow hardware traces of a method executing on a processor configured to generate hardware tracing information. The controller may generate a profile based on the hardware tracing information and initiate re-compiling or re-optimizing of the method in response to determining that the new profile differs from the previous profile. The controller may repeatedly profile and re-optimize a method until profiles for the method stabilize. Profiling and hardware tracing of an application may be selectively enabled or disabled allowing the controller to respond to later phase changes in application execution by re-optimizing, thereby potentially improving overall application performance.

Abstract: Embodiments of the present disclosure relate to a method, a device, and a computer program product for managing scheduling of a backup task. A method for managing scheduling of a backup task is provided, including: acquiring a scheduling setting for a target backup task, the scheduling setting specifying at least a start time of executing the target backup task in a backup system; acquiring execution information associated with execution of the target backup task according to the scheduling setting; and determining a score for the scheduling setting based on the execution information, the score indicating a degree of impact of the execution of the target backup task according to the scheduling setting on the performance of the backup system. With the embodiments of the present disclosure, the scheduling setting can be evaluated, thereby optimizing the scheduling of a backup task and improving the performance of a backup system.

Abstract: A system is provided for optimizing deployment of a processing function in a media production workflow. The system includes a media production workflow generator that builds the media production workflow that includes the processing function and determines deployment criteria that includes an input dataset for the processing function and an atomic compute function for executing the processing function. Moreover, a deployment topology generator generates a topologies of the resources available in a cloud computing network and based on the determined deployment criteria, with the generated topologies indicating different configurations of resources for executing the processing function and a processor for executing the atomic compute function of the processing function.

Abstract: An inventory of computer resources requiring backup data protection including an estimate of future growth of storage resources along with a set of data retention policies, rules, and outcomes in conformance with business requirements, is created. An architecture for initial ingest of protected data and a backup execution approach is created, based on the resource inventory, a sizing and assumptions of resources and data, and the set of data retention policies, rules, and outcomes. An initial ingest copies the protected data to the backup storage. The initial ingest architecture is transformed to a steady-state ingest phase, by incrementally adjusting the resources based on progress of the initial ingest. Responsive to completing the initial ingest of protected data, a steady-state backup ingest is performed. A validity of the sizing and assumptions of architecture resources is determined and an adjustment to the resources is performed based on the validity determination.

Abstract: A data processing system for providing a black-box optimization exchange implements receiving a black-box optimization problem, presenting the problem on a user interface, receiving a candidate solution, inserting the candidate solution into a candidate solution queue, providing the candidate solution to a simulator service cause the simulator service to execute an objective function associated with the black-box optimization problem on the candidate solution, determining an improvement value provided by the candidate solution by analyzing one or more previously determined solutions to the first black-box optimization problem and the first objective value to determine an amount of improvement provided by the candidate solution, determining a reward by analyzing the improvement value using a reward function associated with the black-box optimization problem, and notifying the user that provided the candidate solution that the candidate solution has earned the user the reward.

Abstract: A method for establishing system resource prediction and resource management model through multi-layer correlations is provided. The method builds an estimation model by analyzing the relationship between a main application workload, resource usage of the main application, and resource usage of sub-application resources and prepares in advance the specific resources to meet future requirements. This multi-layer analysis, prediction, and management method is different from the prior arts, which only focus on single-level estimation and resource deployment. The present invention can utilize more interactive relationships at different layers to effectively perform predictions, thereby achieving the advantage of reducing hidden resource management costs when operating application services.

Abstract: In a computing environment, a set of executing processes each having associated resources are provided. Aggregate resources for the computing environment include multiple different types of resources. A utilization level for each of the resources within the computing environment is evaluated to determine an unconsumed capacity for each of the resources below a utilization threshold. The utilization threshold is resource-dependent. An indication of at least a portion of unconsumed capacity for each of the resources below the utilization threshold is gathered. The unconsumed portion for each of the resources below the utilization threshold is exposed for consumption by other executing processes.

Abstract: A system and method for optimized operation of real-time control applications.

Abstract: A method and system for controlling granularity of transaction recording and visualizing system performance and behavior in a discrete functional verification software simulation environment is disclosed. According to one embodiment, a simulation of a model is run in a discrete event simulation system for a period of time. During the simulation, statistical values of attribute for a plurality of transactions occurring during the period of time are monitored. Based on a granularity setting, a group of consecutive transactions is grouped into a super transaction, and the statistical values representing the super transaction are recorded to represent the group of transactions. The super transactions are visualized in a visualization tool for analyzing the performance of the model.

Abstract: An apparatus, system, and method are disclosed for defining normal usage of a computing system resource. A method for defining normal usage of a computing system resource includes receiving a repeating schedule that represents system usage of one or more computing resources and receiving one or more demand events that will affect the system usage of the one or more computer resources. The method also automatically creates a predictive temporal profile that represents the system usage of the one or more computer resources from information comprising the repeating schedule and the one or more demand events. The predictive temporal profile is displayed for the user to review.

Abstract: Systems and methods are provided for detecting an anomaly in a computer that is part of a population of networked computers. Snapshots are received from a plurality of computers within the population of computers, where individual snapshots include a state of assets and runtime processes of a respective computer. An asset normalization model is generated from the snapshots and serves as a baseline model for detecting an anomaly in the state of assets and runtime processes of a respective computer. A snapshot from at least one of the computers is compared to the asset normalization model in order to determine whether an anomaly is present in a state of static assets and runtime processes of the at least one of the computers.

Abstract: A tool for detecting potential impairment in an electronic device. The tool classifies the user input into various categories and determines some measurement, typically a rate of input per user or per time, for determining average use statistics of the electronic device. The tool then determines if a calculated measurement based on user inputs deviates from the average use statistics past some threshold level of deviation. In response to surpassing the threshold level of deviation, the tool identifies the electronic device as potentially impaired.

Abstract: A method of isolating relevant components associated with a performance problem. The method comprises creating a static model of a system under test, including determining relationships between components in a middleware system. The method also comprises creating a dynamic problem tree by applying the static model to the system at runtime. The method further comprises identifying a link probability for each relationship in the dynamic problem tree; and traversing the dynamic problem tree to detect redundant components. A component is redundant if it has a link probability with its downstream component greater than a pre-determined threshold value. The method also comprises removing redundant components from the dynamic problem tree; and returning a set of components which are potential components to be analyzed further to determine the root causes of the performance problem.

Abstract: A test system uses an instrumented browser to identify events that were not successfully captured during a client web session. The identified events can be used to modify a capture system that captures the client web session. Alternatively, the test system may generate replay rules that are used by a replay system to infer the missed events while replaying of the previously captured client web session. The events can include changes to Document Object Models (DOMs) for web pages used during the web sessions. The DOMs can be used to identify significant web session events and force replay sessions into the correct states.

Abstract: In a system operations management apparatus, a burden to a system administrator when providing a decision criterion in detection of a failure in the future is reduced. The system operations management apparatus 1 includes a performance information accumulation unit 12, a model generation unit 30 and an analysis unit 31. The performance information accumulation unit 12 stores performance information including a plurality of types of performance values in a system in time series. The model generation unit 30 generates a correlation model including one or more correlations between the different types of performance values stored in the performance information accumulation unit 12 for each of a plurality of periods having one of a plurality of attributes.

Abstract: In the system management server, an information processing apparatus that is an event-information acquisition target is registered as a monitored apparatus in configuration information; event information that complies with a rule stored in advance is identified from among a plurality of pieces of event information stored in the system management server; a server apparatus for a network service related to the event information is identified; and a message is displayed which indicates that the cause of the event that occurred in a client information processing apparatus which has generated event information is an event related to the network service, which occurred in the server apparatus.

Abstract: Techniques for performing diagnostics are described. In one embodiment, in response to an alert or a request to perform diagnostics, a topological relationship is generated comprising a set of applications and a set of systems determined based upon information in the alert or request. The topological relationship encapsulates relationships between the set of applications and the set of systems. In one embodiment, a set of causal network models to be used for performing the diagnostics is determined based upon the applications and systems in the topological relationship.

Abstract: The illustrative embodiments disclose a computer implemented method, apparatus, and computer program product for refining statistical profile data. In one embodiment, the process interrupts an execution of a software program in response to detecting a profiling event during the execution of the software program. The process then records an event sample associated with the profiling event. The event sample is recorded in a log file. The process also inserts a set of markers describing and associated with the event sample into the log file. The process then resumes the execution of the software program.

Abstract: Provided is a calculating apparatus that calculates a characteristic of a target signal, comprising a designating section that receives a designation of either a bit error rate or a sampling timing; and a calculating section that calculates a range of sampling timings over which the bit error rate is less than a designated value or a bit error rate at a designated sampling timing by using a relational expression between the sampling timing and the bit error rate in a transmission model for transmitting a signal having jitter that includes a random component and a deterministic component having a prescribed probability density distribution, the relational expression achieved by substituting, as parameters, a standard deviation of a random component in jitter of the target signal and a peak-to-peak value of a deterministic component in the jitter of the target signal.

Abstract: In order to investigate the behavior of a data storage system, the time differences between commands from an analysis device and the completion of the operations defined by those commands are analyzed. The method may use a factorial parameter table definition representing common threads of I/O events in a patterned, cyclic nature, using a host computer as the host application emulator and data logger. The workload definition is provided either from hypothetical proposed thread parameters, or by statistical and common factorial analysis of real application I/O traffic analysis, or a combination of both. The generated workloads can be run while measuring the I/O event completion timings and event queuing characteristics, facilitating engineering interpretation of results for the purpose of analyzing the effects of workload and system architectural design factors, and optimization or evaluation of total system configurations.

Abstract: A system and method for prioritizing alerts includes extracting invariants to determine a stable set of models for determining relationships among monitored system data. Equivalent thresholds for a plurality of rules are computed using an invariant network developed by extracting the invariants. For a given time window, a set of alerts are received from a system being monitored. A measurement value of the alerts is compared with a vector of equivalent thresholds, and the set of alerts is ranked.

Abstract: An engine for performing root cause analysis on a system comprising a plurality of elements in a modeled domain, the engine comprising a module loader and parser for loading modules of computer code containing computer language statements modeling elements of the system, including one or more of model type definitions, conditions, causal rules, and continuations; and for parsing the computer language statements; a storage for storing the parsed computer language statements; a receiver for receiving conditions representing state changes of system elements, accessing the storage, and determining one or more parsed computer language statements to apply to the received conditions; a hypothesis generator for instantiating one or more hypotheses of potential causes of each received condition consistent with the received conditions and the parsed computer language statements; a blackboard for receiving the instantiated hypotheses, and determining a root cause of the received conditions based on the hypotheses consis

Abstract: A method and apparatus for fault analysis and fault isolation in a system of networked processors by using a central event correlation function and logical fault signature to provide for fault isolation of failed processing elements is presented. This central event correlation method uses asynchronous events from multiple input sources of same and different technologies and time-based fault correlation and ageing to match unique fault signatures and determine levels of fault recovery escalation over time. This mechanism uses an event driven recovery table to recognize a unique fault signature, count and age faults, provide fault threshold based recovery and generate events as needed to drive recovery escalation.

Abstract: Systems and methods are disclosed that profile event sequences by creating a mixture model from the event sequences; estimating parameters for the mixture model; and applying the mixture model to profile the event sequences.

Abstract: In accordance with a particular embodiment of the present invention, a method is offered that includes generating an automatic test case generation using model checking for web applications, the automatic test case generation including: developing a specification; verifying a property using model checking on the specification; obtaining a counterexample, whereby the counterexample is mapped to a web test case; and executing the web test case on an implementation. In more specific embodiments, the method includes generating counterexamples by negating a desirable property and then model checking the specification, whereby the counterexamples represent a set of witnesses that are mapped to the web test case; and executing the web test case on the implementation. In still other specific embodiments, the generating step and the executing step are repeated on available properties and on their available counterexamples. The witnesses can be mapped to the web test case through selected framework technology.

Abstract: A system such as a Web-based system in which a plurality of computers interact with each other is monitored to detect online an anomaly. Transactions of a service provided by each of a plurality of computers to another computer are collected, a matrix of correlations between nodes in the system is calculated from the transactions, and a feature vector representing anode activity balance is obtained from the matrix. The feature vector is monitored using a probability model to detect a transition to an anomalous state.

Abstract: Techniques are provided that generate bit reliabilities for a detected sequence. A detector generates the detected sequence. According to one embodiment, a post-processor finds a first set of combinations of one or more error events in the detected sequence satisfying a complete set or a subset of error correction constraints corresponding to the first bit value, finds a second set of combinations of one or more error events in the detected sequence satisfying a complete set or a subset of error correction constraints corresponding to the second bit value, selects a first most likely combination of one or more events of the first set and a second most likely combination of one or more events of the second set, and generates a bit reliability based on the first and the second most likely values.

Abstract: A system adapted for analyzing performance and scalability of at least one computer system is disclosed. The system include a monitoring module configured for capturing at least one metric of an application for analyzing performance and scalability and a performance knowledge repository created and represented by using the at least one metric captured by the monitoring module. The system further includes a diagnosis module adapted for mapping the at least one metric to the performance knowledge repository using a diagnosis module and an analysis module configured for conducting performance and scalability analysis based on the mapping of the at least one metric for detecting at least one undesirable condition.

Abstract: A method and system for identifying a configuration parameter of a “sick” computer system that is at fault for causing an undesired behavior based on analysis of configuration parameters from other computer systems is provided. In one embodiment, a troubleshooting system collects “suspect” values for “suspect” configuration parameters used by a “sick” application when the undesired behavior was exhibited by the sick computer system. The troubleshooting system then compares the suspect values to sample values of the suspect configuration parameters retrieved from sample computer systems. The troubleshooting system uses that comparison to identify one or more suspect configuration parameters that are likely at fault for causing the application to exhibit the undesired behavior.