Apama and StreamBase unearth meaning in disjointed data streams
Competing ESP solutions help you navigate rushing rivers of information
In simpler times, early stream-monitoring apps took months to crank out and didn’t easily scale for reuse outside their initial scope.
Today two top-tier products, Progress Apama 2.4 from Progress Software and StreamBase 3.5 from StreamBase Systems, are altering the ESP (event stream processing) landscape. By improving operational insight and automated response — in finance, health care, or general SOA infrastructure monitoring, for example — both solutions can be used to quickly develop apps for building up event correlation out of a flood of enterprise data.
Compared with slower traditional BAM (business activity monitoring), ESP allows you to pull historic data from back-office sources; but its forte is working on data in real time without delays. It offers additional power by detecting missed events, scanning for what didn’t happen, rather than just reporting on what already did.
Although their aims are the same, these two products take decidedly different tacks. The Progress solution inherited a rules-driven approach from the company’s acquisition of Apama: It uses state diagrams and rules composition to generate native applications. In contrast, StreamBase’s Java platform employs visual flow diagramming and its own StreamSQL — a SQL derivative amended for live, time-centric queries on event streams.
Click for larger view.
Apama’s capabilities do not extend much beyond the product’s original trading focus, and development is not quite as easy as with StreamBase or other rules-centric products such as AptSoft Director for CEP. It offers no code editor similar to the one found in StreamBase or iSpheres EPL Studio. Admittedly, the push today is to hide complexity, but there are times when direct access is just easier.
Apama’s capabilities, however, are rich, and its applications aggregate and process data streams well. Included tools — such as the Dashboard Wizard for easy client creation and additional packages for simulation testing and distributed administration — go a long way toward building up the product’s enterprise credibility.
Neither solution can import from external modeling tools, nor does either offer higher-level impact analysis like Tibco BusinessEvents. Both need to improve areas such as integrated process monitoring, both at run time and during simulation. But this remains an emerging field, and vendors are still grappling with approaches and a lack of standards.
The essential system components from both vendors consist of a core runtime engine for processing and correlating events in memory, and a visual IDE used to piece together apps.