Analyzing routing failures

Guidelines: When analyzing routing failures, follow these guidelines:
  • Make sure that the address used in attempting to contact the remote host is a valid IP address.
  • Make sure routing is possible in both directions. For most TCP/IP communication, two-way routing is required. The origin must have routes to reach the destination, and the destination must have routes to reach the origin. If NETSTAT ROUTE at the origin shows correct routing, you must also use NETSTAT ROUTE at the destination to verify that it can send replies back to the origin. If there are intermediate hops between the source and destination, all routing tables must have routing information. For example, if the origin node routing table indicates that the first hop to reach the destination is router A, then the router A routing table must also have a valid, active route to the destination, and so on. This also applies to the return route.
  • Also, this is affected by SOURCEVIPA. If SOURCEVIPA is enabled at the origin of the communication, then the destination and all intermediate hops must be able to route back to the VIPA.
  • If the NETSTAT ROUTE output on the source, the destination, or an intermediate hop does not show the expected routes, do one or more of the following:
    • Make sure that the routers involved in providing routing information are operational and participating in the correct routing protocol.
    • Make sure that the necessary physical connections are active.
    • Use the OMPROUTE DISPLAY commands described in z/OS Communications Server: IP System Administrator's Commands to determine whether anything in the configuration or current state of OMPROUTE has caused the unexpected NETSTAT ROUTE information.
Parent topic: Routing failures