Each morning I drive down out of the mountains of eastern San Diego County to the coastal plains (where the wooden spaceship is docked). This morning I was treated to a scene of sublime beauty — a nearly full moon setting into the Pacific Ocean, through patchy clouds whose edges were backlit by lunar rays. Gorgeous!
Less sublimely beautiful (but perhaps more useful in your day job) are Discovery IP Ranges — the way that you tell Discovery what parts of your network to explore. The first screenshot at above right shows the default IP Address Range type; this may be the only type of Discovery IP Range you've used. To create a range of this type, you simply enter a starting IP address and an ending IP address. Discovery will explore every IP address between the two, including the starting and ending IP addresses. There is one thing you'll need to be careful of: make certain that you do not include any network or broadcast addresses within this range — this would cause very confusing results! For example, suppose your organization's networks are divided into class "C" networks (with 256 addresses each). If you entered a range of 10.10.10.200 to 10.10.11.10, that range would include a broadcast address (10.10.10.255) and a network address (10.10.11.0). This is a problem because when Discovery does something with a broadcast address (such as ping), then all the devices on that network will answer. Many devices will do the same thing with a network address.
The second screenshot shows an IP Network type. This is a very convenient way to specify IP ranges if your organization uses regular (i.e., CIDR-compatible) subnetting — and most organizations these days do. Such networks are commonly written like this: 10.10.10.0/24, where the IP address part before the slash is the network address, and the integer after the slash is the number of bits in the network mask. I used this IP network in my example. When you use an IP Network type, Discovery automatically takes into account the network address and the broadcast address, and excludes them from any exploration — so this example has exactly the same results as my first example. We recommend you use the IP Network type if your organization uses regular subnets, because they will help you avoid accidentally including broadcast and network addresses.
The last type is the IP Address List. This type of Discovery IP Range is most useful when you have devices at well-known, fixed IP addresses that you want to keep refreshed in your CMDB. For instance, you might have 8 network printers, all on static IP addresses, at a particular location. Rather than explore a large range of addresses, perhaps you'd rather manually maintain a list of the eight particular IP addresses you want Discovery to explore. The downside of doing this is that Discovery will not automatically find new devices added to the network. On the other hand, if you have something like a field office connected only via an expensive ISDN link, it might be worth the hand maintenance of the IP Address List in order to cut down the unnecessary traffic over that WAN link. For situations like that, the IP Address List is your friend.
