Hyperic and managing by exception

A short while ago, I had the opportunity to speak with the good folks of Hyperic once again. The conversation originally centered on the topic of their new release, Hyperic HQ 3.2, but then went off on a very important tangent, managing resource by exception. Let's start with a quick description of the new software release.

Quick Review of Hyperic HQ 3.2

Here's a summary of how Hyperic describes the new version of Hyperic HQ.

Hyperic Inc. announced the new version of its flagship product, Hyperic HQ 3.2. Capable of monitoring 1.5 million metrics per minute using MySQL as a backend database, the new software provides operations teams worldwide with the most scalable web infrastructure management solution available.

Hyperic 3.2 adds several new capabilities that help reduce the workload of operations teams charged with managing high-growth web environments. New features include widespread scalability enhancements; support for MySQL as a backend database to Hyperic; cross-platform diagnostic tools that dramatically reduce the steps needed to resolve performance problems; and a new feature set that allows Nagios administrators to automatically integrate and extend their current deployments using Hyperic.

Snapshot Analysis

Hyperic is the business front end of a dynamic open source management software project that has often been one of the most active projects hosted on SourceForge.net. What this means to you is that the developers of this software work for organizations that are actually using this technology. These developers are bringing real-world requirements and experiences and working with others of like mind. They've developed a set of tools that is powerful and quite adaptable to the processes and procedures of the organization.

By moving from their own data store to the popular MySQL database, they've not only improved the scalability and performance of the management environment but, they've also opened the door for others to add new capabilities to the management environment or embed the management environment inside of their own products.

Management by exception

Many management frameworks have been available and in use inside of organizations of all sizes for literally decades. They've grown in power and in levels of available control to the point that they can overwhelm an IT administrative team with detail. Many of the suppliers of these products seem to have forgotten that what these IT administrative teams are trying to accomplish is to make sure that their IT infrastructure is reliable and performs well. Although it might be nice to be able to look at several billion discrete management events that ocurred during the last week, administrators are too busy to enjoy that exercise very much.

A better approach is to allow the IT administrative team to set basic policies of what should be done when certain events occur and what to do if something else that is beyond the basic policy criteria occurs. The management software should then do its level best to create an environment that a former colleague at IDC, Paul Mason, used to describe as "self-managing and self-healing."

After my long discussion with the folks at Hyperic, I came to the conclusion that they really understand that concept and have worked with their open source community to fulfill that need.