i-scream CMS homepage


The i-scream central monitoring system provides a free, centrally configurable distributed server to monitor the status and performance of Solaris, Linux, FreeBSD and Windows machines on a network. The server can raise alerts using email, IRC or a large public information display. Information may also be browsed interactively via the web or using our multi-platform client.

Follow the link above for the original problem specification we were given to develop the i-scream CMS, and the feature set of the CMS after it was finished.

Screen Shots

These screenshots were taken at the end of the project, and therefore show the old build given below, and none of our new features.

Reports Demo

This is a demo of the web-based reports that the i-scream system can do. It's taken from the old builds, so doesn't show any of the new RRDTool based features.


Snapshot Builds

Since the project was completed various changes have happened in the CMS. These are undocumented, but fairly straightforward. We'd strongly suggest you use this version of the CMS rather than the builds from the end of the project.

We've done a fair amount of work on the graphing/database side of the CMS, and fixed numerous bugs. Also, the new ihost/libstatgrab combination requires the snapshot builds.

Old Builds (from project completion)

These builds can be found in the cms directory on our download mirrors. They were completed at the end of the project itself, and match with the documentation below, and work. But we've done a fair amount of work since creating them.


Product documentation

This documentation was written as the CMS project was completed. It therefore applies to the downloads available from the time the project finished. There will definately be descrepencies between this documentation and the snapshot builds.

Javadoc documentation

Each component of the CMS written in java has fairly extensive javadoc documentation that helps to explain the code and API's.

Process documentation

Various pieces of documentation were generated during the project that were intended for internal use. They're not required reading for running the CMS, but might be interesting if you're curious about the project itself.

Project History

The i-scream Distributed Central Monitoring System started off as a final year project for a group of Computer Science students at the University of Kent, UK. The project was completed at 4pm on 29 March 2001. It consisted of five lever-arch folders containing mostly source code and documentation. The CVS repository was tagged to signify this event, however.

As can be seen in our "problem domain" document, the project was aimed at the UKC Computer Science department, and as such we believed that a demonstration to the "customer" was important. The feedback from our demonstation was more positive than we had anticipated. We believe that the system may now be of interest to anybody who wishes to monitor a number of machines on their network using a central configuration. The i-scream central monitoring system is now being used by the UKC Computer Science department, and their feedback is being used to further improve the product.

The name i-scream arose from a meeting whereupon one member of the development team felt a strange urge to draw an ice cream on the discussion board. It later became apparent that we had no name for our project (other than 'The Project'). It seemed only natural to name our project after the ice cream that had been oh-so-randomly drawn on the white board. Copyright issues immediately ruled out anything involving Mr Whippy, so we had to settle for i-scream. And thus, we were left with a cunning play on words - it still sounded like ice cream, yet also served as a functional description of our system - namely, one which screams at somebody if something has gone wrong...