Restartable subsystems

Configuring VMCF and TNF as restartable subsystems has the following advantages:

Error detection is provided when the subsystems do not seem to be initializing properly.
You can change the system name on the restart.
Commands are available to remove users from internal tables, display current users and to terminate the subsystem.

In summary, a restartable VMCF and TNF configuration provides better availability and is therefore recommended.

If you choose to use restartable VMCF and TNF, follow these steps:

Update your IEFSSNxx member in SYS1.PARMLIB with the TNF and VMCF subsystem statements required by TCP/IP. The specification can be in either the IBM® recommended keyword parameter form or the positional parameter form of IEFSSNxx. For example:
```
* The keyword parameter form is:

SUBSYS SUBNAME(TNF)
SUBSYS SUBNAME(VMCF)

* The positional parameter form is: 
TNF
VMCF
```
Add procedure EZAZSSI to your system PROCLIB. A sample of this procedure is located in the SEZAINST library.
```
//EZAZSSI PROC P=&SYSNAME.
//STARTVT EXEC PGM=EZAZSSI,PARM=&P,TIME=1440
```
Start VMCF and TNF using the procedure EZAZSSI before starting TCP/IP. If your node name is the same as the MVS™ system symbolic &SYSNAME, then you can start VMCF and TNF with the following command:
```
S EZAZSSI
```
If your node name is different than the MVS system symbolic &SYSNAME, start VMCF and TNF with the following command:
```
S EZAZSSI,P=nodename
```
Replace nodename with the system name of your MVS system. If you use SMTP, verify that the SMTP NJE node name is specified by using the NJENODENAME statement in the SMTP configuration data set.