START_SHDI_FIRST stop during upgrade ECC 6.0 SR1

33 Views

Hello guys

Hello

We are doing upgrade from R/3 4.6C to ECC 6.0 SR1, platform iseries

In phase START_SHDI_FIRST we get error at starting shadow instance.

It seems the problem is caused by a bad call, the user is DESadm (!!??)

*****************************************************************+

*USRSPC.

Deleting spooled files of instance user

DES03...

0 archivos suprimidos. 0 archivos no

suprimidos.

Starting instance 03 of system

DES...

El subsistema R3_03 ya está

activo.

Command for starting system Job is SBMJOB CMD(CALL PGM

(R3INLPGM)

PARM

('/usr/sap/put/exe/sapstartsrv'

'pf=/usr/sap/put/DES/SYS/profile/START_DVEBMGS03_IBM550' '-

D'))

JOB(SAPSTRSRV) JOBD(R3DES400/R3_03) JOBQ(QSYS/QUSRNOMAX) USER

(DESadm)

SYSLIBL(SYSVAL) CURLIB(CRTDFT) INLLIBL( R3DES400 QTEMP

QGPL)

SRTSEQ(HEX) CCSID(500) CPYENVVAR(YES) ALWMLTTHD

(*YES)

No se ha encontrado el perfil de usuario

DESADM.

Errores producidos en el mandato

SBMJOB.

StartInstance: Error submitting sapstartsrv

job.

***************************************************************+

So, as the user DESadm doesn't exist, the process cancels.

We have applied patch for STARTSAP, but we get same error

Afterwards we have tried to apply patch for R3INLPGM unsuccessfully,

because we got a unknown error during APYR3FIX

Then, we have used a workaround: We have created user DESADM, copied

from DESOFR, and adding additionnal authorizations. With this solution

we don't get anymore this error, of course, and Instance Shadows

starts, but after 2 o 3 mins all workprocess die. Looking the trace

file we don't see the cause ...

So, I think the main problem is a bad call at time of starting shadow

instance.

Any idea???

Thanks and regards

Javier