TCP/IPv6 Migration Guide

Procedures for Migrating to NonStop TCP/IPv6
HP NonStop TCP/IPv6 Migration Guide—524524-004
4-2
Migrating Your Application to NonStop TCP/IPv6
Migrating Your Application to NonStop
TCP/IPv6
1. Find out if TCP6MON is running in all processors.
->STATUS MON $ZZTCP.*
2. Decide whether to use logical network partitioning (LNP). If you do not use LNP, be
aware that all applications share access to all IP addresses in the system. This
condition is similar to Parallel Library TCP/IP but different from NonStop TCP/IP. If
you do use LNP, applications can be restricted to using specific IP addresses by
selecting a TCP6SAM process that is associated with that IP address. This
condition is different from Parallel Library TCP/IP but similar to NonStop TCP/IP.
See Logical Network Partitioning (LNP) on page 2-2 and TCP/IPv6 Configuration
and Management Manual for more information.
3. Decide whether to use round-robin filtering. If you do use round-robin filtering,
determine the key and add the DEFINE. See the TCP/IPv6 Configuration and
Management Manual.
4. Determine the name of the TCP6SAM process for use as the transport-service
provider. See Determining the Name of the Socket Access Method on page 2-1.
5. Add the DEFINE, set the PARAM, or issue the procedure call to set the
transport-service provider to the TCP6SAM process name. See Guardian Socket
Applications on page 2-5 and OSS Socket Applications on page 2-5.
6. To use IPv6 sockets, you must port your socket applications to IPv6. See the
TCP/IP Programming Manual for procedures for porting applications to IPv6.
Note. If your application can be spawned in any processor, and you do not configure a
TCP6MON in every available processor, you need to change your application so that it does
not spawn to a processor without a TCP6MON object.