Steps for NCPROUTE connection problems

Diagnose NCPROUTE connection problems.

Procedure

Perform the following steps:

  1. For an NCP client, make sure that the internet interfaces (token ring and Ethernet) and NCST logical units for communication with the TCP/IP host are defined correctly in an NCP generation. See the ACF/NCP IP Router Planning and Installation Guide for detailed information about NCP definitions.
    1. Make sure that the NCPROUTE UDP port (UDPPORT keyword), coded on the IPOWNER statement in an NCP generation, matches the value defined in the .ETC.SERVICES data set. If it is not coded, the value used is the default UDP port 580.
    2. Verify that the assigned port numbers and service names for NCPROUTE and the router are correct. Also make sure that the router service port 520 is defined in the .ETC.SERVICES data set. The NCP clients use this port as a destination port when broadcasting RIP packets to adjacent routers.
    3. Make sure that NCST logical units for the SNALINK LU0s are defined correctly. A partner LU name (INTFACE keyword) for the SNALINK-NCST interface, coded on the LU statement in an NCST GROUP of an NCP generation, should match the LU name in a SNALINK LU0 DEVICE statement in the hlq.PROFILE.TCPIP data set.
    4. Make sure that the remote LU name (REMLU keyword) for the SNALINK-NCST interface, coded on the LU statement in an NCP generation, matches the VTAM® application name in the VTAM APPL definitions for SNALINK LU0s. For more information about SNALINK configuration and VTAM APPL definitions, see z/OS Communications Server: IP Configuration Guide.
    5. Make sure that the NCST partner LU name (INTFACE keyword) for the SNALINK-NCST interface, coded on the IPOWNER and IPLOCAL statements in an NCP generation, matches the partner LU name in Step 1b.
    6. Make sure that the IP address for the TCP/IP host (HOSTADDR keyword), coded on the IPOWNER statement in an NCP generation, matches the IP address for the SNALINK LU0 device name coded on the HOME statement in the hlq.PROFILE.TCPIP data set.
    7. Make sure that the IP address for the SNALINK-NCST interface (LADDR keyword), coded on an IPLOCAL statement in an NCP generation, matches the IP address for the SNALINK LU0 link name coded on the Start of changeBEGINROUTESEnd of change statement in the hlq.PROFILE.TCPIP data set.
    8. Make sure that the destination IP address for the SNALINK-NCST interface (P2PDEST keyword), coded on an IPLOCAL statement in an NCP generation, matches the IP address on the IPOWNER statement in Step 1e.
    9. Make sure that IPLOCAL statements are defined for the directly-attached NCP internet interfaces (token ring and Ethernet) in an NCP generation. Verify the correctness of the IP addresses (LADDR keyword), metric values (METRIC keyword), protocol type (PROTOCOL keyword), and subnetwork masks (SNETMASK keyword).
  2. Make sure that the appropriate NCP LOADLIB is used and that it contains correct network definitions. The NCP LOADLIB must be in the search list referred to by the //DD STEPLIB statement. Verify that a 374x communications controller to be in the session with NCPROUTE is loaded with the correct NCP load module.
  3. Make sure that appropriate cataloged procedures for NCPROUTE (NCPROUT) and SNALINK (SNALPROC) are used, and verify the correctness of the data set references.
    • For the SNALINK cataloged procedure, make sure that the number of SNALINK sessions is large enough to allow multiple NCP sessions with NCPROUTE. This number is referred to by the MAXSESS keyword on the EXEC statement.
  4. If using SNMP, make sure that the appropriate cataloged procedure for the SNMP agent (SNMPD) is used and verify the correctness of the data set references. Do likewise for an SNMP client (SNMPQE on MVS™ host).
  5. Make sure that NCPROUTE is configured correctly in the hlq.PROFILE.TCPIP data set. The cataloged procedure name (NCPROUT) is referred to on AUTOLOG (optional), and PORT statements. UDP port 580 must be reserved for NCPROUTE.
  6. Make sure that NCPROUTE is configured correctly in the ETC.SERVICES data set. See also Step 1a.
  7. Make sure that SNALINK LU0 is configured correctly in the hlq.PROFILE.TCPIP data set. The SNALINK device name, LU name, and VTAM application address space name are referred to on the DEVICE statement. The SNALINK link name is referred to on the LINK, HOME, and BEGINROUTES statements. See also Steps 1b, 1c, 1e, and 1f.
    • If more than one NCP client is to be in session with NCPROUTE, repeat Step 7 to configure SNALINK LU0 for another session. TCP/IP definitions must be defined for each SNALINK LU0 session. If TCP/IP is currently running and another NCP client is to be added, another SNALINK LU0 can be configured using VARY TCPIP,,OBEYFILE commands. This allows TCP/IP to be reconfigured without having to shut down TCP/IP.
  8. If you are using OMPROUTE, make sure that the routing parameters in the OMPROUTE configuration file (network interface definitions) and TCP/IP configuration (BSDROUTINGPARMS and BEGINROUTES statements) for the NCP clients are defined correctly. In addition, verify that direct and static routes to the NCP clients are defined correctly in TCP/IP BEGINROUTES statement.
  9. If you are using SNMP, make sure that the SNMP agent is configured correctly in the hlq.PROFILE.TCPIP data set. If the SNMP client is on an MVS host, verify that the SNMP client address space is also configured. The cataloged procedure names, SNMPD, for the SNMP agent and client, are referred to on the AUTOLOG (optional), and PORT statements.
    • For the SNMP agent, make sure that the access authority information is defined correctly in the SEZAINST(EZBNRPRF) data set for the NCPROUTE profile, referenced in the NCPROUTE cataloged procedure.
  10. If an NCP client is activated and ready to establish a session with NCPROUTE, make sure that the cataloged procedures for TCPIP, NCPROUTE, and SNALINK are all started. If you are using SNMP, make sure that the SNMP agent and client are started.
    1. Make sure that the SNALINK devices are started by the START statement in the hlq.PROFILE.TCPIP data set. The SNALINK devices can also be started by a VARY TCPIP,,OBEYFILE command or a VARY TCPIP,,START command.
    2. Make sure that VTAM command prompts at the system operator console are replied to; otherwise, a SNALINK session can be in a pending activation state.
    3. Make sure that the NCP client physical and logical lines for the internet interfaces (token ring and Ethernet) are active.
    4. Make sure that NCST lines are active for the SNALINK LU0 sessions.
    5. Make sure that VTAM cross-domain resource managers (CDRMs) are active in the MVS hosts.
  11. For network connectivity problems, see Diagnosing network connectivity problems.
  12. For SNMP problems, see Diagnosing Simple Network Management Protocol (SNMP) problems.
  13. For OMPROUTE problems, see Diagnosing OMPROUTE problems. Ensure that the interface definitions in the BSDROUTINGPARMS statement in the hlq.PROFILE.TCPIP data set match the definitions in the corresponding interface definitions in the OMPROUTE configuration file. For more information about defining BSDROUTINGPARMS for NCPROUTE, see z/OS Communications Server: IP Configuration Reference.