Automatic network routing (ANR) is a low-level routing mechanism that minimizes cycles and storage requirements for routing packets through intermediate nodes.

An ANR node is an intermediate network node on the path of an RTP connection. ANR nodes are not aware of SNA sessions or RTP connections passing through the node. All an ANR node must do is read the header in a network layer packet and forward the information to the next node on the path. The ANR information is learned by the RTP endpoints during establishment of the RTP connection by sending a "Route Setup" message which flows through all nodes on the prospective HPR path.

The ANR functions and services include:

• Fast packet switching

  ANR takes place at a lower layer than APPN intermediate session routing and significantly improves performance in the intermediate nodes. The ANR node routes the HPR packet and does not provide functions such as link-level error recovery, segmentation, flow control, and congestion control. These functions are performed at the RTP connection endpoints.

• No session awareness

  Intermediate nodes that implement ANR are not aware of the SNA sessions or the RTP connections that are established across the nodes. This means that there is no requirement to keep the routing tables for session connectors that are kept in basic APPN.

• Source routing

  Source routing is a technique whereby the sender of the data specifies the route that the data should take through the network. In an IP network (which is a connectionless network), every router examines the data's header and selects the next hop. In an ANR network, the end node or the network node selects the route and the network nodes along the route forward the data based on selection of the originator.

ANR is a source-routing protocol and carries the routing information for each packet in a network header with the packet. Each node strips off the information it has used in the packet header before forwarding onto the link, so the next node can easily find its routing information at a fixed place in the header. This means that switching packets through a node can be done more quickly than in the routing table lookup method used in basic APPN. There is no restriction on the number of hops in ANR.

Figure 1 shows how ANR routes messages.

Figure 1. ANR routing

In the figure, the intermediate network node receives the message and strips the first routing label (A1) from the message header before forwarding the packet on link A1. The address of C5 represents the endpoint in the last HPR node. Since the routing is predetermined, the intermediate network node routes the packets very quickly, with no need to examine the request unit (RU) and/or the transmission header (TH), reserve storage or buffers, or to do link-level error recovery. The same mechanism takes place on the reverse route.