The NSS server can be started in the following ways:
- Using an MVS™ procedure from
the MVS operator console. A sample
start procedure is provided in SEZAINST(NSSD).
- From the z/OS® shell, by
starting OMVS and then issuing the nssd command.
- Using the COMMNDxx member of PARMLIB.
This allows the NSS server to be automatically started when the system
is IPLed. For information about the use and configuration of the COMMNDxx member
of PARMLIB, see z/OS MVS Initialization and Tuning Reference.
- Using the AUTOLOG statement in the TCP/IP profile.
Tips: - You should not start the NSS server using the AUTOLOG statement
in a stack's profile. If the NSS server is listed in a stack's AUTOLOG
statement, the server is cancelled if it is already running when that
stack starts. This results in the NSS server losing any cached information
that it has in place for NSS clients previously connected through
all stacks, and could increase the overall recovery time when a TCP/IP
stack recycles.
- If you start the NSS server from the z/OS shell and you stop the shell environment
from scrolling, then when the nssd command needs
to display data to the shell, the NSS server might stop and wait indefinitely
for the shell to scroll and make output buffer space available for
the data.
- When running from an MVS procedure,
set the environment variables using the STDENV DD statement in the
NSS server procedure.
Restriction: Only one instance of the NSS server can run
on a z/OS image. If you attempt
to start a second instance, the NSS server will fail.