A Natural Resource Company

The natural resource company has sites across North America. You choose to manage the entire network centrally with a single, large NNM system located in the corporate data center. This UNIX system is intended to support 30 concurrent users, so you install the maximum number of CPUs and the maximum amount of RAM supported. You install a second, small backup NNM system at another location.

Rather than depend on snmpCollect-style SNMP polling for performance management, you install RMON probes at all critical LAN and WAN points. HP NetMetrix manages the probes and downloads the performance data. The probes also generate threshold alarms. This removes the performance overhead of collecting performance data from the NNM system and offers better control over the data collected.

The IT staff took all the formal HP OpenView training and hired a consultant for two full weeks to install NNM, direct the discovery of the network, set up SNMP data collection, train the staff, buy lunches, and document the entire network management strategy and implementation.

No discovery filter is used. Instead, you manually guide the NNM discovery process by managing subnets as they are discovered until all of the network is on the map. Here, you depend on a map builder who knows the corporate network topology and can partition the Internet submap intelligently. As each subnet is managed and discovery in it takes place, you observe what devices are found and then unmanage and hide those you don’t want to manage. This usually means hiding workstations and non-SNMP devices and managing network infrastructure, printers, and servers.