|
Data source integration can be one of the most time consuming tasks in developing any event based solution. Some estimates rate data integration as high as 80% of the overall effort. Once events appear at the doorstep of the event processing engine, configuring rules for the remainder of the solution is relatively easy. While event processing engine technology has continued to mature, applications still remain narrowly focused within small segments of the enterprise and little attention has been placed on the “last mile” from the event source to rules engine. At Event Zero, we believe that a framework to address the last mile is an essential component required to enable event processing solutions across the enterprise. Our highly scalable Event Processing Network is a fully integrated solution that includes the framework necessary to rapidly integrate data from many types of heterogeneous sources, distributed in multiple geographic locations, and across a large number of endpoints.
|
Historically, event processing has been used for applications in environments where there is a convenient source of event data and the need to process only a small number stream of events. Typically the event data is acquired by monitoring traffic on an enterprise message bus (JMS, ESB, MOM, etc.) and then sent to an event processing engine on the same server or network for analysis. Event acquisition is relatively straightforward due to a small number of centralized endpoint event streams, although the number of events may scale quite large.
As a result, most event processing technology today has evolved to fit this design pattern where most of the emphasis is placed on the event stream processing engine. This arrangement works well for a limited set of problems that involve analyzing streams of similar yet unrelated events/transactions for trends (ex. Financial trading). However, in order to solve a broader range of business problems with event processing, additional platform functionality is required.
|
Event Zero and the Last Mile
| At Event Zero, we take a holistic approach to event processing for the enterprise with emphasis placed on the following design goals associated features: |
- Rapid integration of new data sources
- Adapters:
- protocol interface modules used to acquire data from an endpoint. Over 60 adapters
exist today and others can be easily created using the adapter API.
- Profiles : define data acquisition, transformation, and enrichment rules required to create an EPN event from an endpoint adapter
|
Large numbers of endpoints
- Agents : a fully managed EPN component used to acquire data from one or more endpoints using one or more types of adapters. Agent software is self updating and may be deployed directly on an endpoint, intermediate server, or other EPN node.
|
Large numbers of events
- Edge processing : the ability offload event acquisition to the outer edges of the EPN
- Event processing pipeline : the ability to distribute the event processing pipeline elements to many servers/EPN Nodes.
|
Many geographic locations
- Edge collectors : an EPN node that collects and processes events in a geographic location before transporting on for further rule processing
- Wan support : the ability to encrypt and compress blocks of events for optimized transfer via a WAN
- Distributed rule engines : ability to host many instances of event processing rules engines throughout the EPN to provide hierarchical event analysis
|
Easy to manage
- Distributed configuration management : ability to remotely configure and manage any EPN node or component
- Self monitoring : the entire EPN, individual nodes, and components are self monitored for health and performance
|
|