What you will learn in Event Processing in Action
Chapter by chapter
Chapter 1: Chapter 1. Entering the world of event processing
This introduction defines event processing broadly and notes the confusion surrounding the term.
What you learn in chapter 1 →Chapter 2: Chapter 2. Principles of event processing
This introduction previews the chapter’s treatment of event-based programming, event processing concepts, and their modeling.
What you learn in chapter 2 →Chapter 3: Chapter 3. Defining the events
This introduction positions event-type definition as a starting point for application design and previews the chapter’s treatment of events.
What you learn in chapter 3 →Chapter 4: Chapter 4. Producing the events
This introduction distinguishes event producers, event processing agents, and event consumers by their roles in an event processing system.
What you learn in chapter 4 →Chapter 5: Chapter 5. Consuming the events
The introduction frames information abundance as a challenge for limited attention and previews the chapter on event consumers.
What you learn in chapter 5 →Chapter 6: Chapter 6. The event processing network
This introduction describes the event processing network as a model of the processing between event producers and consumers.
What you learn in chapter 6 →Chapter 7: Chapter 7. Putting events in context
The section illustrates how everyday circumstances affect situations and decisions, then introduces context as a building block for event processing.
What you learn in chapter 7 →Chapter 8: Chapter 8. Filtering and transformation
This introduction previews filtering, transformation, and the application of context to event processing agents.
What you learn in chapter 8 →Chapter 9: Chapter 9. Detecting event patterns
Introduces event patterns as relationships among events that can reveal significance missed when events are considered individually.
What you learn in chapter 9 →Chapter 10: Chapter 10. Engineering and implementation considerations
The introduction explains how distributing event processing across processors can help avoid centralized-server bottlenecks.
What you learn in chapter 10 →Chapter 11: Chapter 11. Today’s event processing challenges
The chapter introduction previews unresolved challenges involving time, uncertainty, event retraction, and causality in event processing.
What you learn in chapter 11 →Chapter 12: Chapter 12. Emerging directions of event processing
This introduction frames the chapter as the authors’ personal reflections and predictions about emerging directions in event processing.
What you learn in chapter 12 →Concepts explained in this book
- context partitionAn independently tracked portion of an event-processing context that groups events according to inclusion criteria and maintains its own processing and state.
- event processingA computing approach in which events are processed through operations such as filtering, transformation, and pattern detection.
- input terminalA named endpoint on an event-processing component that receives events from other entities.
- event orderingThe sequence in which events are arranged or processed, which can affect order-sensitive results.
- event contextAn event context is a named specification of conditions that groups event instances into partitions for related processing.
- event processing applicationA component that performs a specific function in an event processing system.
- filter expressionAn expression that determines whether an event instance is selected or excluded, or cannot be evaluated.
- translate agentA translate agent is a stateless transformation agent that derives one output event from one input event using a derivation formula.
- composite contextA context formed by combining two or more contexts, with partitions formed from combinations of its members’ partitions.
- segmentation-oriented contextAn event context that groups event instances into partitions according to event-attribute values or predicates over those attributes.
- derivation formulaA rule or expression that assigns values to the attributes of a derived event, potentially using input-event attributes and other data.
- event processing networkAn abstract model describing the functional behavior of an event processing system without necessarily representing its physical realization.
- derived eventA derived event is an event produced by an event processing system from detected information about a situation.
- occurrence timeOccurrence time is the timestamp or interval representing when an event occurred in the external system.
- enrich agentAn enrich agent is a translate agent that adds information to an event using data obtained from a global state element.
- project agentA project agent is a translate agent that derives an event containing a subset of the input event’s attributes.
- detection timeDetection time is the timestamp at which an event became known to the event processing system.
- temporal contextA temporal context is a context defined by one or more time intervals that determine which events are considered together.
- participant event setThe set of event instances within a pattern detect agent's context partition that have types listed as relevant to the pattern.
- event processing systemA software system that processes representations of events to detect situations or produce observations about them.
- header attributeA metadata field associated with an event instance, whose value may be fixed by its event type or vary from instance to instance.
- compose agentA transformation agent that combines two input event streams by matching event instances and deriving output events from matching pairs.
- aggregate agentAn event-processing agent that groups input events into context partitions and derives output events from them.
- event typeA specification for a set of event objects that share semantic intent and structure.
- output terminalA named endpoint on an event-processing component through which it emits events.
- event-driven architectureA way of designing and operating applications that use events directly or indirectly.
- relevant event types listA list specifying which event types qualify as participants and how many instances of each type are required by a pattern.
- static analysisStatic analysis is the examination of a program or expression without executing it to identify properties or errors.
- split agentA split agent derives a collection of events from one input event, with each output containing all or some of the input attributes.
- clock synchronization problemThe challenge of aligning event occurrence timestamps produced by clocks that may not agree.
- order-sensitive processingEvent processing whose result or behavior depends on the order in which events are processed.
- retrospective compensationA technique for addressing late events by retracting affected results and replaying events in their correct order.
- timeout reorderingA method for handling out-of-order events by buffering them for a bounded delay and sorting them by occurrence time.
- event channelA communication path through which events are conveyed between event-processing components.
- XPathA language for selecting and addressing parts of an XML document.
- filter agentAn event processing agent that tests incoming events to decide whether to discard them or pass them on.
- event streamAn event stream is an open-ended sequence of associated events arriving over time.
- data typeA classification that specifies the kind of value an attribute can hold.
- event processing patternA recurring form of event behavior or event relationships that can be detected by an event processing system.
- event retractionThe withdrawal or invalidation of a previously reported or derived event and its effects.
- matching setA matching set is a collection of event instances that satisfy a matching criterion.
- graphical notationGraphical notation is a visual convention for representing components and connections in an event processing network.
- explicit channelAn event channel explicitly represented in a model, potentially with identifiers and configuration parameters.
- threshold assertionA pattern parameterization consisting of a threshold relation and a threshold value, used to test an aggregation result.
- event consumer nodeA modeling construct that represents an abstract event consumer type, a class of concrete consumer instances, or a single concrete consumer instance, and describes its interaction with the rest of an event processing network.
- stateless event processingA transformation that processes events without depending on or maintaining state between them.
- event processing building blockAn abstract modeling construct representing an event-processing concept independently of its implementation, from which definition elements can be created.
- event transformationEvent transformation is an operation that changes an event or its representation.
- reference dataReference data is information that event-processing agents can consult during processing but that the event-processing system does not maintain.
- temporal granularityThe smallest unit of time, or chronon, that determines timestamp precision for an event type.
- stateful event processingStateful event processing is processing in which handling events is influenced by more than one input event.
- all patternAn event processing pattern satisfied when the participant event set contains at least one instance of every listed event type.
- loose couplingDecoupling is the independence of event-producing and event-consuming applications, which need not know of one another.
- continuous queryA query that remains active to process incoming data continuously rather than returning a result from a fixed input.
- pattern detection agentAn event-processing agent that examines incoming events for specified patterns or relationships.
- pattern typeA label that determines an event pattern's meaning and intention and specifies the kind of matching function to use.
- scalabilityScalability is a system’s capability to adapt to changes in use, volume, or demand while continuing to meet its business objectives.
- pattern signatureA specification of a pattern that includes its pattern type, relevant event types, and, optionally, parameters and policies.
- threshold patternA pattern that compares an aggregation of participant events with a threshold value using a specified relation.
- spatial contextA context that groups events by spatial conditions without temporal conditions.
- definition elementA platform-independent modeling construct that represents an event type and is an instance of a basic building block.
- state-oriented contextA context that groups event instances according to the value of a state, such as weather conditions.
- event certaintyEvent certainty is an estimate of how likely it is that a particular event occurred as described by its payload.
- event instance headerAn event instance header is the part of an event instance containing identifying and descriptive attributes about the event.
- any patternAn event processing pattern satisfied when the participant event set contains an instance of any listed event type.
- parallelismThe execution of separate parts of a computation simultaneously across multiple processing units.
- stream-oriented languageStream-oriented programming is a style organized around processing streams of data.
- joinAn operation that matches records or events from two inputs according to a condition and combines matching pairs.
- recoverabilityRecoverability is the ability to restore a system's state after a failure.
- agent instanceA runtime instance of an event processing agent that performs its function for a particular set of events or data.
- runtime checkRuntime validation is the checking of data or program results against requirements while a system is executing.
- open contentOpen content is free-format data that an event instance may carry in addition to attributes defined by its event type.
- consumer classA consumer definition element representing a class of event consumers rather than a single instance.
- count patternAn event pattern that detects when the number of qualifying events reaches a specified threshold.
- value average patternA threshold pattern that tests the arithmetic mean of a specified numeric attribute across participant events.
- value max patternA threshold pattern that tests the maximum value of a specified attribute across participant events.
- value min patternA threshold pattern that tests the minimum value of a specified numeric attribute across participant events.
- asynchronous programmingProcessing in which work can proceed without requiring each operation to wait for the previous one to finish.
- event interval contextA context type whose window is defined by an initiator event and an expiration condition.
- minimum distance patternAn absolute distance pattern satisfied when the minimum distance of participant events from a point satisfies a threshold assertion.
- spatial patternA pattern satisfied according to distances between event locations.
- matchingThe comparison of event information against criteria or patterns to determine whether they correspond.
- event processing platformAn event processing platform is software infrastructure that provides facilities for developing, executing, distributing, and managing event-processing logic.
- raw eventAn event introduced into an event processing system by an event producer.
- derivationA computation that determines output event content from input event content.
- flat event structureAn event type whose attributes each occur once and have simple data types, without nested or variable structure.
- payload complexityThe degree to which an event payload supports structures, repeating or optional attributes, and open content.
- load balancingThe distribution or adjustment of work among available computing resources to manage workload.
- partitioningThe grouping of event processing agent instances into runtime artifacts for execution.
- data flow graphA directed graph whose nodes represent processing elements and whose edges represent data flow between them.
- data streamA stream is a potentially unbounded, continuous flow of data items.
- specializationA relationship in which one event type inherits another event type's definition, with possible additions or modifications, and represents a narrower category.
- timestampA timestamp is a data type denoting a particular point in time, with precision determined by temporal granularity.
- event type identifierAn attribute that uniquely identifies the event type described by a definition element.
- context initiator policyA semantic abstraction that defines what happens when a new initiator event is detected while a context partition is active.
- distance patternA pattern that evaluates distances between event locations against a distance condition.
- relative min distance patternA relative distance pattern satisfied when the minimum distance between any two participant events satisfies a threshold assertion.
- business process management systemA business process management system is software for modeling, executing, and managing business processes.
- distributed processing modelThe execution of processing work across multiple networked machines.
- epa assignment optimizationAn optimization approach that maps event processing agents to physical runtime artifacts to influence system performance.
- inexact processingInexact event processing accommodates imprecision in event information and interpretation.
- consumption policyA configuration that determines whether an event instance may participate in subsequent matching sets after being included in one.
- api-based mechanismA mechanism that distributes events through a programming interface provided by an event processing network.
- protocol-based mechanismA mechanism that distributes events using a protocol supported by an event processing network implementation.
- dynamic analysisAnalysis of a program or system based on its execution and observed behavior.
- repeated type policyA semantic rule defining how a pattern-detecting agent behaves when it encounters more instances of an event type than specified.
- white box optimizationA white box optimization changes the implementation of a component rather than treating it as fixed.
- rule-oriented languageA programming language organized around expressing and evaluating rules.
- not-selected terminalAn output point through which participant events not included in any matching set are emitted when a temporal window closes.
- edgeAn edge is a connection between vertices in a graph, and in a directed graph it has an assigned direction.
- fixed location contextA type of event context that groups events according to their locations relative to predefined spatial entities.
- spatial relationshipA geometric relationship used to compare an event location with an entity location.
- pattern policyA pattern policy is a setting that tunes the semantics of an event pattern when finding related subsets of events.
- stateful filteringFiltering in which the decision about an event can depend on accumulated state or other events.
- sequence patternAn event processing pattern satisfied when instances of listed event types occur in the specified order.
- performance objectiveA specified target for a system’s performance, expressed through one or more metrics.
- event headerThe portion of an event that contains metadata attributes, distinct from its payload.
- adapterA software component that enables interaction between systems with incompatible interfaces.
- runtime artifactA software module that can be an event processing engine, an engine instance responsible for a specific function, or code implementing a specific function.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.
