Version 2 (modified by smith@…, 8 years ago) (diff)

First attempt at overview, still need to describe ICA/MUA/Web

Raptor Overview / Understanding Raptor

Raptor is a free to use, open source software suite generally designed to enable accounting within event-based systems, and specifically aimed at accounting within systems that handle authentication events such as the Shibboleth IdP, OpenAthens LA, and EZproxy.

The academic IT services community is the user community with the requirements that led to the development of this software, but it should be more generally applicable.

Raptor Architecture and Components

Raptor is designed with a recursive hierarchical architecture:

  • Raptor ICAs (Information Collector Agents) sit on the systems to the monitored for events to be accounted for. Once gathered, this event information is sent to a Raptor MUA.
  • Raptor MUAs (Multi Unit Aggregators) collect event information sent to them and store it. Raptor MUAs can also send events onwards to another Raptor MUA.
  • RaptorWeb is a Web interface that talks to a Raptor MUA, allowing users to view the event information in summary or queryable form.

Raptor Architecture Diagram Raptor Architecture (Diag and explanation)

Raptor ICA (Information Collector Agent)


Raptor MUA (Multi Unit Aggregator)


Raptor Web (Web Interface to a Raptor MUA)


Hierarchical Aggregation


Raptor Deployment

The recommended deployment strategy for Raptor (see here for installation instructions) is to install the Raptor Server components (MUA/Web) onto a dedicated Raptor server within an organisation, and deploy the Raptor ICA on each system that you wish to monitor.

However, for those organisations with limited resources and who only have a single existing service hosted on a single server they wish to monitor accounting data for, all of the Raptor components (ICA/MUA/Web) can be installed together co-located with the service itself (see here for installation instructions).