We have explored the possibility of integrating NWS in the proposed DIT.

We propose a conservative extension of the original DIT, that should be appropriate also for those tools that offer the management of a set of sensors, and are not based on a logfile located on the sensor host.

We provide a prototype of an information provider, to verify the feasibility and the efficiency of the approach.

We analyze the performance of NWS in a specific environment, and observe that queries take a significant time to complete. We suspect that this is not due to communication delays, and therefore a careful location of the processes might not improve significantly the performance.

One indication for a next revision of is that queries should be run in parallel whenever possible. Otherwise the number of sensors, and the refresh period must be balanced.

...after all, NWS is not that much...

Previous: Performance issues

WP7 meeting 24-1-2002