The chapter "Extensible SNMP Agents" in Mauro and Schmidt's Essential SNMP begins, "There will come a time when you want to extend an [SNMP] agent's functionality." So-called "extensible" agents can not only extend the values in a MIB, but can obtain near real time values from external systems as if they'd been returned by the agent.But has the full capability of a knowledge-based system been exploited in the SNMP context? It's known, for example, that Tivoli embeds a rule-based framework. Mauro & Schmidt survey OpenView, Net-SNMP and SystemEDGE agents, which they consider representative of the extensible agent framework. The book doesn't attempt to cover this topic deeply, which can be quite complex if the goal is to build situation awareness about a network's overall computing environment. Add a visualization component to this, and you've got one of the main challenges facing the large network management packages.
Except for essentially academic experimentation, it seems clear that much more can be done in this area than has been attempted so far.
◦
Except for essentially academic experimentation, it seems clear that much more can be done in this area than has been attempted so far.
0 comments:
Post a Comment