|
I got through the first part of Kevin Leonard's gen process. I did run into
some issues, though...
1) The biggest issue is that the starter system only has 4 3350s defined. I
discovered this was a problem while getting g04-init-dasd-misc to run. I
wound up unmounting the WORK61 volume at 14B and using that for PERM73
instead.
2) The starter system needs to be IPLd from a 3215 console at 01F. Thai is
because that address needs to be online at IPL time so the SPECIFY HARDCOPY
message can be replied to with "r 00,'hardcpy=01f,all,cmds)'". Replying
SYSLOG instead of 01F there results in the SPECIFY HARDCOPY message being
repeated. You can have a 3270 console connected at 010, and once the system
is up, switch to it with "v 010,mstcons".
3) The supplied stage 1 has display consoles defined, and so a 6.7 genned
with that deck can't be IPLd: you have to have the 3270 online at IPL time
or else you get a 00F wait state, and if you have it online the DIDOCS bug
gets you. I'm about to rerun the stage 1 with the 3270 consoles removed.
Maybe the usermod can be applied before the gen is run? Haven't looked at
that yet.
--
Jay Maynard
|
Hi.
Somehow sorry to resurect an old post... I missed this one and created one which, essentialy is an ask for help because I'm stuck on your first issue in your list.
I'm quite confused on the actual setup of the starting system regarding the 4 DASDs initialization.
Reading your post makes me feel that maybe those notes I found on the .cmd file were, indeed, outdated or misleading, because I don't fugure out what is the picture at that point:
Do you remember which was the arrangement (hercules .conf CUUs -> files and/or hercules attach CUU -> file commands, and the vary/mount commands?
Cheers!
|
Jay:
You're right on all counts: the starter system has a very limited set of UCBs, and the documentation for using the starter system is inadequate. I'm inclined to leave the starter system as much as possible the way IBM built it, so people can have a fairly genuine experience of using the starter system, but the peculiarities need to be better documented. I don't know if the VS1 6.7 3270 console problem is present in 6.0 -- I don't think it is, but I'm not sure -- but without an actual hardcopy console device, IPLing with a graphic console would require SYSLOG, and I don't think the starter system supports SYSLOG, so we're stuck with the printer/keyboard console IPL. We could document switching to a graphic console after IPL as you note. The documentation definitely needs work. Thanks for your thoughts, and let us know what else you run into. I hope at some point to start working on building a more permanent "turnkey" VS1 system, and definitely want to make the document more robust as part of that process.
--
Kevin
|
Hi folks.
Being an absolute noob... I went through the process of building 6.0.
It was extensively documented, pdf, screenshots...impossible to fail... and when I saw the 6.7 README as the sole reference, I was very excited to give that a try, expecting issues and doubts (that's how I learn the most :-) and using the 6.0 experience to 'fill the gaps'... but I failed to figure out through the UCBs.
I also was keen to try sysgenning 6.7 because, although I somehow managed to IPL 6.0 under VM/370 (Inspired by a Moshix Youtube channel video) , I felt the whole thing was missing stuff, and I though that 6.7, with those instructions to include compilers, etc were the way to go.
So far, I never (AFAIK) experienced any issue regarding 3270s on 6.0 (maybe because I ran it atop VM).
Anyways, thank you very much for your help... this is very nice!
|
Hi again folks!
Thanks to the points you gave me last week, I managed to make through (more or less) to the point on the README where I'm suposed to IPL the new generated system (FMGEN67, which I keep at 148), perform the first time a r 00,'auto=cold' and then, apparently, keep on sending the remaining jobs...
The thing is that, while I don't have any clue of some step having gone wrong until now, once I have IPLd 148 and the format is done, the resulting IPLd system doesn't seem to be friendly either... I'm unable to send any job with this system, I feel its kind of incomplete.
More specifically, I miss any message indicating a job entry started on any partition, no IEF403I INITSWA STARTED or IEF005I PARTITION WAITING FOR WORK messages. In fact, a whole IPL handshake now is quite brief (and discouraging):
HHC01018I 0:001F COMM: client 10.112.10.61 devtype 3215: connected
IEA775I REAL STORAGE LIMITED TO 8192K
IEA760A SPECIFY VIRTUAL STORAGE SIZE
IEA103I DATASET SYS2.LINKLIB NOT FOUND BY LOCATE
IEA130I LINK LIBRARY DATA SETS NOT FOUND
SYS2.LINKLIB
IEA761I PAGE=(V=PAGE73,BLK=8192)
IEE054I DATE=95.177,CLOCK=14.44.20
IEE054I DATE=95.177,CLOCK=14.44.20,GMT
IEA764I NIP01,CMD01,DFN01,JES01,SET01,SMF01
IEA765I DEVSTAT=ALL
IEA101A SPECIFY SYSTEM AND/OR SET PARAMETERS FOR RELEASE 06.7 OS/VS1
IEA103I DATASET SYS1.DUMP NOT FOUND BY LOCATE
IEA135A SPECIFY SYS1.DUMP TAPE UNIT ADDRESS OR NO
IEA208I SYS1.DUMP FUNCTION INOPERATIVE
IEE140I SYSTEM CONSOLES
CONSOLE/ALT COND AUTH ID AREA ROUTCD
30E/01F H CMDS 05 ALL
010/01F N ALL 01 Z,A ALL
01F/009 M ALL 02 ALL
009/010 N ALL 03 ALL
011/010 N ALL 04 Z,A 1-13,15-16
30E/01F A NONE 05 NONE
IEF032I PARMLIB VALUES TAKEN FOR JES
IEE101A READY
IEE029I JLPRM=(U)
IEF249I FOLLOWING P/R AND RSV VOLUMES ARE MOUNTED
FGEN67 ON 148 (P/R-PRV)
PERM73 ON 14C (P/R-STR)
WORK73 ON 14D (P/R-PUB)
FDLB67 ON 248 (P/R-PRV)
PAGE73 ON 351 (P/R-PRV)
IEE052I MONITOR JOBNAMES,T ** FROM CMD01 AT IPL
IEE009I LOG NOW RECORDING ON DATA SET X
IEE052I MONITOR DSNAME ** FROM CMD01 AT IPL
IEE360I SMF NOW RECORDING ON SYS1.MANX ON 148. TIME=14.44.35
IEE048I INITIALIZATION COMPLETED
I feel as if I'm missing something... at this point on the README, I should execute g12 JCL, where g15 is indeed (as I understand by JCL contents) dealing with SYS2.LINKLIB, however, it seems as if this system does expect SYS2.LINKLIB there... Does someone remember how should one proceed after g11 and/or cold IPLing the new system? I have tried statring initiator but it doesn't work, I'm stuck here.
Thank you very much in advance.
Regards.
|
On Mon, Jun 26, 2023 at 10:10 AM, Alejandro olivan Alvarez wrote:
I managed to make through (more or less) to the point on the README
where are the instructions that you are following? I'm interested in trying VS1.
|
however, it seems as if this system does expect SYS2.LINKLIB there It does. But the fact that it isn't there yet isn't a show stopper. The reason VS1 expects to find SYS2.LINKLIB is that the 'SYS1.PARMLIB(LNKLST00)' member is telling VS1 initialization to look for it: SYS1.LINKLIB,SYS2.LINKLIB The sample generating-system/g10-parmlib.jcl job, which builds PARMLIB members, adds the LNKLST00 member along with several others. Because SYS2.LINKLIB hasn't been created yet, VS1 initialization doesn't find it, and issues the message. After you've run the job to create SYS2.LINKLIB, it will exist and VS1 initialization will subsequently find it, and the message will no longer be issued. You won't need it until then anyway. I suppose I could have created SYS2.LINKLIB before adding the LNKLST00 member, but it seemed more straightforward to do it this way. Maybe it would be appropriate to document the warning messages that will be issued at this point which can be ignored. --
|
On Wed, Jun 28, 2023 at 12:43 AM, Doug Wegscheid wrote:
where are the instructions that you are following? I'm interested in trying
VS1.
Hi!. Within the 6.7 .zip file, there's a folder named 'generating-system', where the different steps jcl jobs sit. There you'll find also a README txt file that has, overall, a kind of 'roadmap' on what to do. It was very helpful to me to first follow the fully-detailed, prior VS1 6.0 generation (different .zip) guide, because, although clearly more limited in scope, it had a pdf document that describe the process (even with screenshots), and it serverd me to get 'the feel' on the VS1 environment and 'fill the gaps' (to some extent) when going through the 6.7 README. As mentioned before, the main issues found are that you've to be very careful regarding a 3270 issue (I got rid of everything except a telnet console), that it is not clear which CUUs to use for the different DASDs/Volumes, and some class issues preventing the jcl decks to be executed... All those issues have been dealt here :-) However, I'm stuck in the process where the 12th JCL has to be run after IPLing the new system... I'm missing something or something went wrong, as decks are no longer read/executed. Cheers!
|
Hi!
Thank you very much for the tip!
However, the issue I'm facing is that the new IPLd VS1 system does not run jobs (or, most probably, and to be precise, I don't know how to manage it to do so).
I've been looking for the IBM Operator guide for VS1, but, although referenced around in the books found on bitsavers for VS1, I have not found it...so I don't know how to perform the very basic operations after IPL.
My guess (out of sheer ignorance) is that, much like DOS / POWER, some partitions and/or JES has to be enabled/activated, because it is not done automatically at this point (I've read around VS1 has some automation).
the display all command shows nothing, beyond 00e and 00d after issuing sf commands.
Sf ,00c command does nothing after devinit a jcl at the reader.
I feel that the system has no partitions active, no JES1, nothing... just itself up...and I don't know what to do.
At this point, the key question is whether this is normal for a system at this stage (indicating that early steps were right and I'm simply missing the knoledge to move on), or it is something unexpected (indicating that something went wrong up to g11.
Regards.
|
My guess (out of sheer ignorance) is that, much like DOS / POWER, some partitions
and/or JES has to be enabled/activated, because it is not done automatically at this
point (I've read around VS1 has some automation). There should be an active partition, if you're using the parmlib members I built. An initiator should have been started in partition P1 automatically by the command: s init.p1 which is in the 'SYS1.PARMLIB(CMD01)' list of automatic commands. A DISPLAY ACTIVE at that point, before you try to start a reader, should show something like: D A IEE821I 18.12.05 ACTIVE DISPLAY 060 C PNO JOBNAME STEPNAME SUBT STAT PNO JOBNAME STEPNAME SUBT STAT WTRZ WTRZ WTRF 00D WTRF 00E 001 INIT P1 A,E Sf ,00c command does nothing after devinit a jcl at the reader. I guess I would have to see it. If the reader is a TCP socket device, a "devlist 00c" command at the Hercules console should show something like: HHC02279I 0:000C 3505 localhost:15741 sockdev ascii eof IO[0] HHC02279I (no one currently connected) If the reader is a file initialized with a "devinit 00c jcl/br14.jcl" command, a "devlist 00c" should show something like: HHC02279I 0:000C 3505 jcl/br14.jcl eof IO[0] before you issue the STARTF command at the VS1 console. It would be helpful if you could post your Hercules log and your VS1 log to give us some idea of where you are. --
|
Hi! Definitely, no active partition. Just for curiosity, I would add that the exact same behaviour is what I got when I built 6.0 using the very detailed pdf that was included. So far, and as far as I'm aware of, I saw no evident error during the steps that did generate both 6.0 and 6.7 systems (the later though, once some issues were overcome). If the resulting systems should automatically bring up P1 (which now I'm thinking it makes sense), they're not doing so (not when done in my environment at least), so something is going wrong. It would be interesting to have a look at what's on SYS1.PARMLIB(CMD01)... however, whithout an active partition, I'm unable to run jobs (and I would struggle to assemble a JCL to do that...I'm still not enough confident handling those JCL / DDs :-P). I'm starting to see some refreshed activity around here, and I have taken note of that syntax command: s init.<PX>[,more stuff] ... I still see some room to play around with this VS1 stuff and have a ton of fun. I have to take a look at some new proposed books around that may help to fill hte gap for the Operators guide we're missing :-) Thank you. Cheers
|
I'm reviewing the listing of the jobs ran to generate the system, and I've found that during the tape copying, amidst a huuuuuuge amount of printing wich looks like trouble free, all of the sudden I stumble with this:
IEB154I BRWPVVC HAS BEEN SUCCESSFULLY LOADED
IEB154I BRWPVVD HAS BEEN SUCCESSFULLY LOADED
IEB144I THERE ARE 0000067 UNUSED TRACKS IN OUTPUT DATA SET REFERENCED BY OA3790MO
IEB149I THERE ARE 0000010 UNUSED DIRECTORY BLOCKS IN OUTPUT DIRECTORY
IEBCOPY MESSAGES AND CONTROL STATEMENTS PAGE 0142
COPY INDD=ISMPACDS,OUTDD=OSMPACDS
IEB167I FOLLOWING MEMBER(S) LOADED FROM INPUT DATA SET REFERENCED BY ISMPACDS -
IEB154I ?.?b?? HAS BEEN SUCCESSFULLY LOADED
IEB154I D?gs?? HAS BEEN SUCCESSFULLY LOADED
IEB154I D?gt ? HAS BEEN SUCCESSFULLY LOADED
Tons of this:
IEB154I G?3? HAS BEEN SUCCESSFULLY LOADED
IEB154I G?\?? HAS BEEN SUCCESSFULLY LOADED
IEB154I G??s?#a HAS BEEN SUCCESSFULLY LOADED
IEB154I G??y?? HAS BEEN SUCCESSFULLY LOADED
IEB154I G???0? HAS BEEN SUCCESSFULLY LOADED
IEB154I G??`}? HAS BEEN SUCCESSFULLY LOADED
IEB154I G???}? HAS BEEN SUCCESSFULLY LOADED
.... And finally
IEB144I THERE ARE 0000017 UNUSED TRACKS IN OUTPUT DATA SET REFERENCED BY OSNAACDS
IEB149I THERE ARE 0000020 UNUSED DIRECTORY BLOCKS IN OUTPUT DIRECTORY
IEB147I END OF JOB -00 WAS HIGHEST SEVERITY CODE
//MOUNT JOB MSGLEVEL=1,PRTY=11
The jobs ends with 00 highest severity code.... but the output for G3 looks really bad :-( ... Is this normal?
Cheers.
|
Yes. SMP uses the CDS/ACDS PDS directory member names to carry information, so they're not printable. On Fri, Jun 30, 2023 at 6:33?AM Alejandro olivan Alvarez < alejandro.olivan.alvarez@...> wrote: I'm reviewing the listing of the jobs ran to generate the system, and I've
found that during the tape copying, amidst a huuuuuuge amount of printing
wich looks like trouble free, all of the sudden I stumble with this:
IEB154I BRWPVVC HAS BEEN SUCCESSFULLY LOADED
IEB154I BRWPVVD HAS BEEN SUCCESSFULLY LOADED
IEB144I THERE ARE 0000067 UNUSED TRACKS IN OUTPUT DATA SET REFERENCED BY
OA3790MO
IEB149I THERE ARE 0000010 UNUSED DIRECTORY BLOCKS IN OUTPUT DIRECTORY
IEBCOPY MESSAGES AND CONTROL
STATEMENTS PAGE 0142
COPY INDD=ISMPACDS,OUTDD=OSMPACDS
IEB167I FOLLOWING MEMBER(S) LOADED FROM INPUT DATA SET REFERENCED BY
ISMPACDS -
IEB154I ?.?b? HAS BEEN SUCCESSFULLY LOADED
IEB154I D?g s? HAS BEEN SUCCESSFULLY LOADED
IEB154I D?g t ? HAS BEEN SUCCESSFULLY LOADED
Tons of this:
IEB154I G?3? HAS BEEN SUCCESSFULLY LOADED
IEB154I G?\?? HAS BEEN SUCCESSFULLY LOADED
IEB154I G??s?#a HAS BEEN SUCCESSFULLY LOADED
IEB154I G??y?? HAS BEEN SUCCESSFULLY LOADED
IEB154I G???0? HAS BEEN SUCCESSFULLY LOADED
IEB154I G??`}? HAS BEEN SUCCESSFULLY LOADED
IEB154I G???}? HAS BEEN SUCCESSFULLY LOADED
.... And finally
IEB144I THERE ARE 0000017 UNUSED TRACKS IN OUTPUT DATA SET REFERENCED BY
OSNAACDS
IEB149I THERE ARE 0000020 UNUSED DIRECTORY BLOCKS IN OUTPUT DIRECTORY
IEB147I END OF JOB -00 WAS HIGHEST SEVERITY CODE
//MOUNT JOB MSGLEVEL=1,PRTY=11
The jobs ends with 00 highest severity code.... but the output for G3
looks really bad :-( ... Is this normal?
Cheers.
-- Jay Maynard
|
On Fri, Jun 30, 2023 at 03:52 AM, Alejandro olivan Alvarez wrote:
It would be interesting to have a look at what's on SYS1.PARMLIB(CMD01)...
however, whithout an active partition, I'm unable to run jobs (and I would
struggle to assemble a JCL to do that...I'm still not enough confident
handling those JCL / DDs :-P).
Alejandro, take a look at ZZSA. It's a standalone system that lets you view and edit PDSs on OS volumes. I have used it on occasion when rendering one of my MVS systems inoperable.
|
On Fri, Jun 30, 2023 at 02:17 PM, Doug Wegscheid wrote:
Alejandro, take a look at ZZSA. It's a standalone system that lets you view
and edit PDSs on OS volumes. I have used it on occasion when rendering one of
my MVS systems inoperable.
mmmm .... missed that site! :-o Thank you for the tip! .... I'm following closely your thread on G13 (I'm thinking on stop posting on this one, since I'm no longer sure the notes I carefully took up to G11 are valid, based on what I got after IPLing). I'll look at the tool you tell me and try to look at the contents of CMD01! Cheers!
|
