Routing Failures
Routing problems are usually the result of outages in a network and a lack of alternative routing paths available for recovery. Routing problems can also be the result of incorrect configurations in the channel-attached and network-attached routers as well as incorrect ARP entries. PING and TRACERTE commands to and from a z/VM® host are useful diagnosis aids for problem determination. If a PING or TRACERTE command fails on a system where MPRoute is being used, a client is unable to get a positive response to a PING or TRACERTE command. Before doing any other problem determination, issue the NETSTAT GATE and SMSG server_id RTTABLE or SMSG server_id RT6TABLE commands on the local and remote hosts to get the routing table information for both the TCP/IP stack and MPRoute.
From the NETSTAT GATE outputs, determine which route is used to reach the destination and determine the route-active state. For IPv4, a routing table is searched in the following order, starting with the most specific to the least specific:
- Host Routes
- Subnet Routes
- Network Routes
- Supernet Routes
- Default Routes
For IPv6, a routing table is searched in the following order, starting with the most specific to the least specific:
- Host Routes
- Prefix Routes
- Default Routes
If there are no active routes available to reach the destination or if there are improperly configured channel-attached or network-attached routers along the routing path, the PING and TRACERTE commands will fail. To function correctly, PING requires active routes in both directions between the PING origin and the PING destination. If the routes are shown to be active at the local and remote hosts, the problem is most likely caused by a router along the routing path. Use the output from the TRACERTE command to locate the suspect router.