Analytic tools are important. In the services arena, making tools accessible is essential for progress in standards, common understanding and sharing of best practice. Hewlett Packard released the ‘Seymour’ services analytics toolkit into the public domain in order to promote standards and information sharing. As the original developers of Seymour we continue to use it and would encourage others to do so.
There are two primary technical considerations for any service - people or technically bounded, performance and availability. Essentially what capacity does our service have to meet (changing) demands and how reliably can it do this over time.
Seymour, available from Sourceforge provides three key tools to help you estimate and manage these
- Demos2K: a process-resource based simulation environment for building models that can help you understand both steady state and ‘surge’ performance of your service based on your resources, human, electronic and physical. Demos has been designed to enable very rapid model prototyping and at the same time have a formal semantics, allowing greater confidence in the validity of results. Demos is documented by Hewlett Packard here, and the modelling approach that it supports is documented here.
- Savant: a validated and flexible fault tree analysis tool that enables rapid exploration of systems architecture reliability, service structures and availability.
- Goldstar: a reward calculation scheme that enables organisations to place performance and availability in the context of their own businesses.