ECC 6.0 EHP4 upgrade to EHP5 fails in MAIN_NEWBAS/XPRAS_AIMMRG

Question

from SAPehpi.log
CURRENTPHASE MAIN_NEWBAS/XPRAS_AIMMRG
...started at 20110825140601
...begin processing at 20110825140601
..times self: 0.080000/0.150000 usr/sys, children 0.560000/0.450000 usr/sys.
..finished at 20110825140820 with status UNRESOLVED REQUESTS IN BUFFER.
Error message set: 'Unresolved requests in buffer UT1
Check logfiles 'XPRASUPG.ELG' and '/usr/sap/putUT1/EHPI/abap/log/SAPehpi.ECO''
...begin dialogue at 20110825140821
from SAPehpi.ECO
EXECUTING /usr/sap/UT1/SYS/exe/run/tp (/usr/sap/UT1/SYS/exe/run/tp) pf=/usr/sap/putUT1/EHPI/abap/bin/XPRASUPG.TPP put
UT1
This is /usr/sap/UT1/SYS/exe/run/tp version 380.03.36 (release 720, unicode enabled)
Warning: Parameter INTERRUPT is no longer used.
Warning: Parameter DAYLIGHT_SHUTDOWN is no longer used.
Warning: Parameter WITH_TACOB is no longer used.
Warning: Parameter IMPDP_BY_EVENT is no longer used.
Warning: Parameter INTERRUPT is no longer used.
Warning: Parameter DAYLIGHT_SHUTDOWN is no longer used.
Warning: Parameter WITH_TACOB is no longer used.
Warning: Parameter IMPDP_BY_EVENT is no longer used.
Warning: Parameter DBCONFPATH is no longer used.
Looking for effective putsteps ... ... ready (looking for putsteps)
ERROR: stopping on error 12 during EXECUTION OF REPORTS AFTER PUT
stopping on error 12 during EXECUTION OF REPORTS AFTER PUT
tp returncode summary:
TOOLS: Highest return code of single steps was: 12
WARNS: Highest tp internal warning was: 0118
tp finished with return code: 12
meaning:
  A tool used by tp aborted
Process with ID 4903 terminated with status 12
from XPRASUPG.ELG
  LIST OF ERRORS AND RETURN CODES  *******
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
XPRA ERRORS and RETURN CODE in SAPR702XPRA00000005.UT1
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1AETR012XProgram terminated (job: "RDDEXECL", no.: "14071100")
  Long text:
    Cause
      Program "&V#&", which was started in the background, was terminated
      abnormally.
    System Response
      The system created a job log for the terminated program.
    What to do
      Proceed as follows:
      Log onto the system in which the program was executed. The system is
      specified at the beginning of the transport log.
      Then call transaction SM37 to display the job log.
      Enter "&V#&" as job name and "DDIC" as user name.
      Restrict the starting date to the starting date specified in the
      transport log.
      For the other selection criteria, select only jobs with the status
      "cancelled". Then press <LS>Execute</>.
      A list of the jobs satisfying the selection criteria is displayed.
      You can display the log by pressing <LS>Job log</>.
      If the list contains several jobs, you can select the job with the ID "
      &V#&" with <LS>Display</> -> <LS>Job details</> or define further
      details for the selection criteria in the initial screen of transaction
      SM37.
        If the ABAP processor reports cancellation, double-clicking on the
      corresponding message in the job log branches to the display of the
      corresponding short dump.
1AEPU320 See job log"RDDEXECL""14071100""UT1"
1 ETP111 exit code           : "12"
>>> PLEASE READ THE REPORT DOCUMENTATION OF THE REPORTS MENTIONED ABOVE <<<
XPRAs are application reports that run at the end of an upgrade.
Most XPRA reports have a report documentation that explains what
the report does and how errors can be corrected.
Call transaction se38 in the SAP system, enter the report name,
select 'Documentation' and click the 'Display' button.
>>> The problematic XPRAs are mentioned in messages of type PU132 above <<<
<end of logs>
I have searched for ST22 logs and i got a syntax error on program  CL_PROXY_STYLESHEET_RUNTIME.
Upgrade 702 has added a segment in the DELETE_CLASS of the program above:
delete from sproxxsl
     where class  = class and method = <sproxxsl>-method and client = <sproxxsl>-client.
before:
delete from sproxxsl
     where class  = class and method.
"client" field is not part of table SPROXXSL. Which is causing the error.
Any idea on how to continue???
The choices now are:
01)  *  Repeat phase MAIN_NEWBAS/XPRAS_AIMMRG to continue at the point it stopped
02)  -  Initialize phase MAIN_NEWBAS/XPRAS_AIMMRG to restart it from the beginning
03)  -  Accept non-severe errors and repeat phase MAIN_NEWBAS/XPRAS_AIMMRG
04)  -  Exit this program
Edited by: Rainier Sison on Aug 29, 2011 2:21 PM
Edited by: Rainier Sison on Aug 29, 2011 2:22 PM
Edited by: Rainier Sison on Aug 29, 2011 2:23 PM

Answer 1

Well, I did have a similar issue with RDDEXECL during the import of SAPKB71007.
The solution at that time was updating the kernel to the latest support level, that was released a few weeks after the error with a special modification to fix 'our' issue. 
Do you think it's possible to change the kernel on the shadow instance at this moment of EHPI execution?
Hope this helps
Best Regards
Francisco
Edit: The OSS note was:          1551400 - ABAP: Session abort when compiling a deep constant
Edited by: Francisco José Martínez Carretero on Aug 29, 2011 4:54 PM