MVS-sysgen/sysgen

SYSGEN fails with 3times RC008 and 53 times RC012

Closed this issue · 9 comments

...
SYSGEN00    IEHPROGM                       0008 <--
SYSGEN01    CLEANUP                        0008 <--
SYSGEN6     STEPZ1                         0008 <--
ZP60008     STEP1                          0000
ZP60008     STEP2                          0000
ZP60008     STEP3                          0000
ZP60008     HMASMP       STEP4             0000
ZP60008     HMASMP       STEP5CK           0012 <--
ZUM0008     HMASMP       RECEIVE           0000
ZUM0008     HMASMP       APPLYCK           0012 <--
...

prt00e_sysgen.2021-08-10-171530.txt.gz

Moshe,

it looks like this is a consequential error from the receive step.

Could you search the output for

HMA4240    HMASMP EXEC PARM = 'DATE=U'
  RECEIVE
          SELECT(ZUM0007)

We have to search for an error at this step.

/Mike

Sure, here it is:

HMA4240    HMASMP EXEC PARM = 'DATE=U'
  RECEIVE
          SELECT(ZUM0007)

HMA3930    SYSMOD ZUM0007 SUCCESSFULLY RECEIVED
HMA4240    HMASMP EXEC PARM = 'DATE=U'
  APPLY
        SELECT(ZUM0007)
        CHECK
        .
HMA3603 ** SMPCDS SYSTEM ENTRY NOT FOUND
HMA2050    APPLY PROCESSING COMPLETED - HIGHEST RETURN CODE IS 12

...

That's looking fine.

So question is, why is the next step failing:

Be on the lookout for stuff like this:

IEB139I I/O ERROR DURING WRITE -         SMPJOB03,SMP     ,148,DA,AOSU0   ,30- OP,PROGRAM CHECK  ,0000007D001621,EXCP
             IEBCOPY MESSAGES AND CONTROL STATEMENTS                               PAGE 0003


IEB101I  I/O ERROR WRITING MEMBER DATA AT TTR=003C21

Has this been resolved?

For what it's worth I've managed to stumble into a setup where it fails pretty consistently.
Naturally no closer to figuring out what is wrong, but it's a start in the right direction.

My plan is to extract that step from the sysgen process, so I can just run it over and over again quickly,
and with different builds of Hercules.

Bill

AND -- I'm going to copy this virtual machine and not use it for anything else.
Last thing I want is for something to change and it stops failing as regularly!

I've created a new version of sysgen which uses python and breaks each step out as its on atomic step which should make bug hunting easier. I'm going to close this issue.