International Business Machines Corporation

Globalfoundries Inc.

Joachim H Frank

Wilfred Cadelina Jamison

Lin Y Yang

Hui Lei

Liangzhao Zeng

Han Chen

Christoph Lingenfelder

E Michael Maximilien

Hung-yang Chang

Kevin Spencer Barker

John William Alcorn

Werner Kriechbaum

Gerhard Stenzel

David Michael Eads

Curtis Reed Miles

James Rumbaugh

Embodiments of the present invention address deficiencies of the art in respect to conversation protocol validation and provide a method, system and computer program product for testing protocol compatibility between trading partners in a B2B conversation. In one embodiment of the invention, a method for testing protocol compatibility between roles in a B2B conversation can be provided. The method can include selecting at least two roles from amongst global process models, constructing local protocols for each of the selected roles, generating a conversation protocol from the constructed local protocols, and detecting dead-end states in the generated conversation protocol to identify protocol incompatibility while identifying protocol compatibility in the absence of dead-end states in the generated conversation protocol.

Testing protocol compatability

A method and associated systems for monitoring and maintaining a cluster of virtual machines. The cluster contains one or more pairs of a first virtual machine and a second virtual machine, in which each machine of a pair monitors the other one machine of the pair. When a first virtual machine identifies that its corresponding second virtual machine is not operating properly, the first virtual machine automatically requests that a system-management entity restart the second machine. If a certain number of restart attempts fails to restore the second machine to desired functionality, the first virtual machine automatically requests that the system-management entity recreate or reprovision the second virtual machine from a prior backup. If a certain number of such attempts fail, a system administrator is automatically notified that further action is needed.

Maintaining a cluster of virtual machines

A method and associated systems for increasing resiliency of a distributed computing system. A processor creates a first virtual machine and a second virtual machine, each of which monitor the other. When the first virtual machine identifies that the second virtual machine has become unavailable or is otherwise compromised, the first virtual machine automatically requests that a system-management entity restart the unavailable machine. If a certain number of restart attempts fails to restore the second virtual machine to desired functionality, the first virtual machine automatically requests that the system-management entity recreate or reprovision the second virtual machine from a prior backup. If a certain number of such recreations or reprovisionings attempts fails, a system administrator is automatically notified that further action is needed.

Increasing resiliency of a distributed computing system through lifeboat monitoring

A virtual machine supports execution of compiler-generated code and provides an execution environment for code that may correspond to any of several different event-driven or message-driven programming models (such as state machines, rules for complex event processing and event correlation, monitoring contexts, processes, and mediations). The programming models may be used in business monitoring and/or business integration environments. A programming language is defined for execution using the virtual machine, which contains a relatively small number of instructions and which is preferably generated by a compiler from a higher-level, event-driven or message-driven model.

Virtual machine and programming language for event processing

A method for non-intrusive event-driven prediction of a metric in a data processing environment is provided in the illustrative embodiments. At least one set of events is observed in the data processing environment, the set of events being generated by several processes executing in the data processing environment. A subset of the set of events are tracked for an observation period, the tracking resulting in bookkeeping information about the subset of events. A pattern of events is detected in the bookkeeping information. The pattern is formed as a tuple representing a process in the several processes, the metric corresponding to the process. A prediction model is selected for the tuple. The prediction model is supplied with the tuple and executed to generate a predicted value of the metric.

Non-intrusive event-driven prediction

A method, system and computer program product for achieving a balance between pre-filter efficiency and pre-filter throughput. Event subscriptions are obtained for a given event consumer. The event subscriptions define the events to be routed to this event consumer. Pre-filters of varying complexity and filtering power are constructed based on these event subscriptions. A balance between pre-filter efficiency and pre-filter throughput may be achieved by selecting the best pre-filter out of these constructed pre-filters to filter events to its associated event consumer. The best pre-filter is selected by comparing a complexity threshold (e.g., a limit for the number of atomic Boolean expressions in a pre-filter) with the actual complexity (e.g., the actual number of atomic Boolean expressions) for each constructed pre-filter and selecting the most complex pre-filter without exceeding the threshold.

Generation of a set of pre-filters from a set of event subscriptions to more efficiently select events of interest

A method, system, and computer usable program product for non-intrusive event-driven prediction of a metric in a data processing environment are provided in the illustrative embodiments. At least one set of events is observed in the data processing environment, the set of events being generated by several processes executing in the data processing environment. A subset of the set of events are tracked for an observation period, the tracking resulting in bookkeeping information about the subset of events. A pattern of events is detected in the bookkeeping information. The pattern is formed as a tuple representing a process in the several processes, the metric corresponding to the process. A prediction model is selected for the tuple. The prediction model is supplied with the tuple and executed to generate a predicted value of the metric.

A compiler generates code providing an execution sequence, from a declarative event-driven model, for execution by a computer. To assist in debugging the declarative event-driven model, the compiler also generates cross-references that relate steps in the generated code to elements of the declarative source code. The cross-references may be used as input to a debugger, along with the declarative source and the generated code, to assist a user who is stepping through the generated code (for example) in understanding which portion of the declarative source gave rise to the current step in the generated execution sequence.

Compiler for a declarative event-driven programming model

Debugger for a declarative event-driven programming model

A model of a process is created using novel 'fan-out' and 'fan-in' symbols. A fan-out symbol represents a point in the process flow where a variable number of homogeneous parallel outgoing threads are being split out from a single incoming thread. The fan-in symbol represents a point in the process flow where a variable number of parallel incoming threads with homogeneous output are combined into one or more outgoing threads.

Modeling homogeneous parallelism

Disclosed are a method and apparatus for processing a continuous audio stream containing human speech in order to locate a particular speech-based transaction in the audio stream, applying both known speaker recognition and speech recognition techniques. Only the utterances of a particular predetermined speaker are transcribed thus providing an index and a summary of the underlying dialogue(s). In a first scenario, an incoming audio stream, e.g. a speech call from outside, is scanned in order to detect audio segments of the predetermined speaker. These audio segments are then indexed and only the indexed segments are transcribed into spoken or written language. In a second scenario, two or more speakers located in one room are using a multi-user speech recognition system (SRS). For each user there exists a different speaker model and optionally a different dictionary or vocabulary of words already known or trained by the speech or voice recognition system.

Growing community of inventors

Greenwich, CT, United States of America

Joachim H Frank