How Observer Infrastructure maps a route

Observer Infrastructure uses ICMP to find routers. This section provides details about how that is done. Understanding this can help you interpret route maps and reports.

Observer Infrastructure uses ICMP echo requests (commonly called pings) to discover the routers and the delay introduced by each router between the host system and the destination address. It does this by using the ICMP TTL (Time To Live) function, which sets the number of router hops that are allowed before the packet is considered stale and then discarded. This prevents packets with unreachable destinations from bouncing between routers indefinitely. Most routers and firewalls respond to pings, but not all do. If a router does not respond to a ping, it is usually configured that way on purpose for security or performance reasons.

By sending out a sequence of ping packets starting with a TTL value of 1, then increasing the value for each subsequent ping, Observer Infrastructure causes each router along the destination path to respond with a time exceeded message.

Because routers can go down and routes can change for other reasons between pings, the accuracy of a single trace of a particular route cannot be guaranteed. For the most accurate picture of a route, it is best to analyze the results of multiple route tracing sessions (which Observer Infrastructure makes easy to do).

In some configurations, the steps along the route between your system and another station never varies, but that is not always the case. This is particularly true when packets are routed across the Internet. Routes may change either because an intermediate router is down or overloaded, or because a group of intermediate routers are load-sharing to optimize connection times.

When polling route maps, Observer Infrastructure records all the routes that it discovers. When multiple routes are discovered, they can be examined by clicking on the Routes tab on the main route map display.