We analized the requirements of a resource monitoring tool, and we designed a way to improve NWS fault-tolerance, scalability and fault tolerance.
As far as we know, our design differs from a similar project that is being carried out by the implementors of NWS.
We implemented a prototype to check the consistency of the design, and to gather experience.
We dealt with the design of an infrastructure allowing the seamless integration of heterogeneous monitoring tools into the same resource monitoring service.
The overall architecture shares some features with the NIMI tool.
The experience gained with this case study was valuable for the
design of the GlueDomains architecture, that we are currently
deploying in a real-scale testbed.
previous: An important design decision