Our system starts under control of an automation product. We have nothing of
significance in the startup commands. They are all issued by the automation
program. We have a "SSID" (Systems Service Identifier) called VMCF like this:
*
* TCPIP Virtual Machine Communication Facility
*
SSID=VMCF
STARTCMD=(Start EZAZSSI,SUB=MSTR)
READYMSG=EZY6011I /* VMCF INITIALIZATION COMPLETE */
READYMSG=EZY6003I /* VMCF Already Active */
. . .
. . .
which has a start command coded into it, which it issues when EZAZSSI needs to
be up. As Skip says, it can then be managed for the life of the system. In this
case, as each subsystem and its dependencies are defined, bringing it down up
or restarting does not need any extra effort or care because the automation
product follows the state of the subsystems and manages them in relation to
each other.
-------------------------------------------------------------------------------------------------------------
Although I know little about TCP/IP, I'd like represent the automation
viewpoint. Although a task can be started in various ways, including (for some)
an option in SSN or (for most) a command in SYS1.PARMLIB, only a task started
by 'automation' (SA, for example), can be managed by automation for the life of
the IPL. Starting a task is the easy part. Restarting it in the case of task
failure or stopping it at system shutdown requires TLC from automation.
In theory an automation product could take ownership of task started outside of
the product, but I have not seen one that does. So if you have an automation
product, be wary of starting a task any other way.
.
.
.
J.O.Skip Robinson
Southern California Edison Company
Electric Dragon Team Paddler
SHARE MVS Program Co-Manager
323-715-0595 Mobile
626-302-7535 Office
[email protected]
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
