The Hercules-MVS 'binary' distribution
Preface
After so any people in the Hercules Community have worked on
MVS 3.8. and many got it up and running there might be some left
that do not have the time (or interest) in building the system 'from
the ground' - but they'd probably like to play with MVS anyway.
Or, there might be people that would like to do there own full
installation - but with a more stable system then the MVS 3.7
starter system.
This distribution contains a fully functional MVS 3.8, build
from the SMP install. It's a pure system with no maintenance
applied (yet). Anyway, it's working quite good. Also VTAM and TSO
is preconfigured. As a bonus for TSO the 'FSE' - a full screen
editor from the CBT tape (http://www.cbttape.org) is also
included. After some very few installation steps, you're ready to
IPL.
In the future it might be easier to supply 'software packages'
(i.e. tools from the cbttape) to some 'standard' distribution ...
I do not want to discourage anyone from building his own
system ... this is just meant to be a little helper. To keep the
download small I've left out the distribution libraries, but I'll
probably publish this one later.
The most current version of this distribution can be found at http://www.schaefernet.de/hercules
- in the future you'll also find corrections and add-ons at that
place.
Have fun with MVS 3.8
If you have questions, comments or suggestions feel free to
contact me at s390@schaefernet.de
- but please don't expect 7x24 support :-))
January 2001, Wolfgang Schäfer
Table of Contents
- Release Notes
- Prerequisites
- Contents of the
distribution
- Installation
- Windows
- Linux
- Your first IPL
- Remarks
Release 1.00 2001-01-23
This is the first release. MVS, JES, TSO and VTAM and FSE
should be fully functional. For the page datasets I've choosen a
3340 disk pack which circumvents a problem with Hercules' DASD
emulation in releases prior to 1.71.
The complete build was done on a Win98 system - on Linux only
a few tests were done.
This distribution assumes that you already have hercules
installed and are familiar with operation of Hercules. For MVS 3.8
you'll also need a 3270 emulator which allows you to open at
least 3 sessions (2 might be o.k. if you remove the secondary
console from the configuration.
The basic disk drives require about 770 MB of disk space. This
includes two (empty) 3330 storage volumes for your datasets.
It is recommended that you unpack the distribution to a
subfolder (this document assumes 'mvs') below your hercules
directory (the one with the hercules executables/binaries). Once
you have unpacked the distribution zip/tgz package you should
find the following files in your destination folder:
- dasd - a subfolder that contains your DASD farm: mvsres,
page01 and sys001 are already there, you'll create some
more DASD during the installation. From the archiv you
should have the following files:
- mvsres - 3350 - your MVS system residence volume
- page01 - 3340 - contains the page datasets
- sys001 - 3330 - used for system installation
libraries, system programmers' stuff and a SMF
dump file, currently the FSE source libs are on
this volume (but not cataloged)
- install - subfolder with installation 'helpers'
- doc - contains this 'manual' + other documentation
- makejobs - contains the jobs I used to create this system
after the SMP accespt. (Stage 1 deck, ...)
If you've unpacked the distribution, you're almost done.
Open a DOS window (Commandline) and change (CD) to the 'install'
subdirectory below your 'mvs' directory.
Note to Win/ME users: there is no DOS window. I could never
use it under ME, but I thing you can create an assoctiation for
the batch files and/or commands so that they use the install
directory as working directory.
At the DOS prompt enter 'create_dasd' - this will execute the
create_dasd.bat file. Using the dasdload program, this batch file
creates 3 3350 volumes: spool01, work01 and tso001. If your 'mvs'
directory is not directly below the folder that contains the
hercules programs, you'll have to modify the .bat file before you
run it.
Next you should review the mvs.cnf file in the 'mvs' directory.
If you have changed any dataset names of the dasd files, you'll
have to do change this file accordingly. You'll also might want
to change the TZOFFSET parameter to match your timezone (if you're
not on CET).
You are ready to IPL!
Change to the install subdirectory of your 'mvs' directory.
Use 'bash create_dasd.sh' - using the dasdload program, this
script creates 3 3350 volumes: spool01, work01 and tso001. If
your 'mvs' directory is not directly below the folder that
contains the hercules programs, you'll have to modify the script
file before you run it.
Next you should review the mvs.cnf file in the 'mvs' directory.
If you have changed any dataset names of the dasd files, you'll
have to do change this file accordingly. You'll also might want
to change the TZOFFSET parameter to match your timezone (if you're
not on CET).
You are ready to IPL!
Time to start MVS for the very first time ... this first IPL
is 'special' since some datasets (i.e. page, spool) are
uninitialized - starting with the next IPL the procedure is
slightly different.
- go to your 'mvs' directory and start Hercules: ..\hercules-370
-f mvs.cnf
- connect three 3270 terminals to port 3270 on your
Hercules machine. The first one (address 010) is going to
be the master console, the second (011) is used as an
alternative console and the third will become a VTAM
terminal that can be used for TSO. If you don't plan to
start VTAM/TSO on this first IPL you can omit the third
connect ...
- on the Hercules control panel enter ipl 150
after a few trace messages the master console
will display the prompt for IPL parameters:
IEA101A SPECIFY SYSTEM PARAMETERS
FOR RELEASE 03.8 .VS2
Reply with r 00,clpa
- this will build the LPA on the first page
dataset. On subsequent IPLs you may
reply r 00,u - bypassing the step to
rebuild the LPA. After a while you'll see two messages:
IEA301I IEALOD00 NOT FOUND
IN SYS1.PARMLIB
IEA891I NO MAXUSER VALUE SPECIFIED, DEFAULT VALUE OF 256
USED
These are informationale messages only and do not prevent
MVS from anything (probably maxuser 256 is a little bit
more then we need ...)
- Now the display switches to the
real console mode and you should see a // START JES2
command, followed by some messages. Two of them need a
reply:
*00 IFB010D ENTER 'IPL
REASON,SUBSYSTEM ID' OR 'U'
IEE354I SMF PARAMETERS
IEE354I SID=H155
IEE354I JWT=10
IEE354I BUF=2000
IEE354I MAN=ALL
IEE354I EXT=YES
IEE354I OPT=2
IEE354I OPI=YES
*01 IEE357A REPLY WITH SMF VALUES OR U
Reply them with r 00,u
and r 01,u (you have to press enter
after each reply ...) The first reply is from SMF asking
you for the IPL reason, it's recorded in the SMF datasets.
The second reply allows you to override SMF parameters -
what we do not want to do at the moment. After a moment
you should see the "SMF NOW RECORDING ..."
message
- This is the time when JES2 starts
up. You'll see a message IEF677I WARNING MESSAGES ... -
you can savely ignore it. The next is the JES prompt for
parameters:
*02 $HASP426 SPECIFY OPTIONS
- HASP-II, VERSION JES2 4.1
. Since we have no 'real' spool, we have to format it:
reply with r 02,format,noreq this will
do a JES2 cold start. On your next IPL you can quickstart
JES by using r 02,warm,noreq
- But not immediately: now the
console display is full. On the right bottom of the
display "MESSAGE WAITING" is shown. For this 'tutorial'
we'll clean the console manually. You can use many
variations of the MVS 'K' command, but we'll use the
simplest form here: just enter K - the
system will print K E,1,9 on the console
command line. Just hit enter to clear the first 9 lines
... (repeat this whenever you like - or your display is
filled).
- Now JES complains:
*03 $HASP479 UNABLE TO
OBTAIN CKPT DATA SET LOCK - IO ERROR -
* REPLY Y OR N TO CONTINUE
reply with r 03,y
- the next prompt will follow very soon:
*04 $HASP436 REPLY Y OR N TO
CONFIRM CHECKPOINT RECORD CHANGE
the reply is r 04,y
(again ...)
Now JES2 formats it's spool.
These two prompts will only occur on the first start,
because the JES datasets (on volume SPOOL1) are new and
not correctly formatted.
- Now JES comes up and your system is
ready for work: $HASP099 ALL
AVAILABLE FUNCTIONS COMPLETE
- Let's start VTAM now: S NET
- a list of messages appears:
- s net
$HASP100 NET ON STCINRDR
- $HASP373 NET STARTED
IST025I BLDL FAILED FOR IEDIAE IN VTAMLIB
IST025I BLDL FAILED FOR IEDIAE IN VTAMLIB
IST025I BLDL FAILED FOR IEDIAE IN VTAMLIB
IST025I BLDL FAILED FOR IEDIAE IN VTAMLIB
IST025I BLDL FAILED FOR IEDIAK IN VTAMLIB
IST025I BLDL FAILED FOR IEDIAK IN VTAMLIB
IST110I NETWORK SOLICITOR STARTED
IST093I APPTSO00 ACTIVE
IST093I LCLNSN00 ACTIVE
IST020I VTAM INITIALIZATION COMPLETE
IEA000I 021,IOE,05,0E00,400000000001,,,,21.14.45
IEA000I 022,IOE,05,0E00,400000000001,,,,21.14.46
I haven't investigated the IST025I yet, but it's
informational ... the IEA000I messages are produced
because VTAM has 3 terminals defined (020, 021, 022) and
only one is connected - so the others will give an I/O
error. But VTAM is working: if you look at the terminal
connected to 020 you should see the "THIS TERMINAL IS LOGGED ON TO THE NETWORK
SOLICITOR" message.
Time to start TSO ...
- Enter S TSO on the
console. Very soon you'll see - IKT005I TCAS IS INITIALIZED, time to switch over to the terminal
connected to VTAM and log on TSO
- This distribution contains two TSO
users: the default IBMUSER and a new one SYSPROG -
SYSPROG does also have all privileges but since he has
"REGION(0)" in his profile he can use more
commands without running out of storage. On your VTAM
terminal enter LOGON IBMUSER now - and *congratulations*
you are logged on to TSO ... LOGOFF again and go back to
the console.
- If JES wishes to print something,
you must start the printer once with $SPRINTER1.
JES will show a message like
*$HASP190 IBMUSER SETUP --
PRINTER1 -- F = STD. -- C = 6 -- T = 0
when it want's to print. The message says that you should
mount the correct form on your printer. Since our printer
is virtual, we do not care ...
- How to bring the system down
- Stop TSO: p tso - TCAS will
prompt for termination options: enter R
nn,U (nn is the reply number)
- After TSO is down, stop
VTAM: z net,quick - at some
point you should see the message "IST102I
VTAM IS NOW INACTIVE" - that may take a
while ... if it's too long, you can force VTAM
out of your system by using z net,cancel -
VTAM will abend, but terminate (you could even
restart)
- Stop JES2: $pjes2 .
The initiators will terminate, some messages
regarding the SYSLOG will appear and finally (again
after a while) the message
$HASP085 JES2 TERMINATION COMPLETE will be on
your master console.
- Now we can 'close' MVS: z
eod. Note that this switches your SMF
file (on the first IPL to SYS1.MANY). And a DUMP
REQUIRED message is shown for SYS1.MANX. After
your next IPL you should dump this file by
issuing S SMFDUMP,MAN=X (and for the following
IPLs or SMF dumps specifing the suffix of the
dataset to be dumped on the MAN= parameter.
Remarks
I know that this documentation is not perfect ... but it's a
start ..., some more will follow ...
Finally a list of the FMIDs I accepted: EAS1102 EBB1102 EBT1102
EDE1102 EDM1102 EDS1102
EER1400 EGA1102 EGS1102 EIP1102 EJE1103 EMF1102 EMI1102 EML1102
EMS1102 EPM1102 EST1102 ESU1102 ESY1400 ETC0108 ETI1106 ETV0108
EUT1102 EVT0108 EXW1102 FBB1221 FDM1133 FDS1122 FDS1133 FDZ1610
FUT1133
...