on 05-10-2016 10:25 AM
Hi all
We hit some issues during the downtime phase of SUM with SPS patching due to a mismatch of physical and virtual host names. This was corrected but SUM crashed, so we had to restart the system.
Unfortunately now we are getting the following error:
SAPup> Starting subprocess with PID 10823 in phase 'TP_ACTION_CP2STRIG' at 20160509163849
ENV: CLASSPATH=:/db2/db2qs0/sqllib/java/db2java.zip:/db2/db2qs0/sqllib/java/runtime.zip:.:/db2/db2qs0/sqllib/java/db2java.zip:/db2/db2qs0/sqllib/java/runtime.zip:.
ENV: DB2DBDFT=QS0
ENV: DB2INSTANCE=db2qs0
ENV: DIR_LIBRARY=/SPS_Upgrade/QS0/SUM/abap/exe
ENV: LD_LIBRARY_PATH=/SPS_Upgrade/QS0/SUM/abap/exe:/usr/sap/QS0/SYS/exe/run:/usr/sap/QS0/SYS/exe/uc/linuxx86_64
ENV: PATH=/SPS_Upgrade/QS0/SUM/abap/exe:/usr/lib64/qt-3.3/bin:/usr/local/bin:/bin:/usr/bin:/usr/sap/QS0/SYS/exe/uc/linuxx86_64:/usr/sap/QS0/SYS/exe/run:/home/qs0adm:.:/db2/db
2qs0/sqllib/bin:/db2/db2qs0/sqllib/adm:/db2/db2qs0/sqllib/misc
ENV: SAPSYSTEMNAME=QS0
ENV: dbms_type=db6
ENV: dbs_db6_schema=SAPQS0
PWD: /SPS_Upgrade/QS0/SUM/abap/tmp
EXECUTING /SPS_Upgrade/QS0/SUM/abap/exe/tp pf=/SPS_Upgrade/QS0/SUM/abap/var/CP2SHDT.TPP r3e SAPKBBO731 -s QS0 u8
This is tp version 380.29.44 (release 721, unicode enabled)
This is /SPS_Upgrade/QS0/SUM/abap/exe/R3trans version 6.24 (release 721 - 01.04.15 - 20:15:04).
unicode enabled version
2EETW000 sap_dext called with msgnr "1":
2EETW000 ---- db call info ----
2EETW000 function: db_ntab (shadow)
2EETW000 fcode: NT_RDTDESCR
2EETW000 tabname: E070~
2EETW000 len (char): 5
2EETW000 key: E070~
2EETW000 retcode: 1
2EETW000 sql error: -10 SQL0010N The string constant beginning with ""XXXXT WITH CS -- OPTLEVEL( 5 ) -- QUERY_DEGREE" does not have an ending string deli
miter. SQLSTATE=42603
2EETW125 SQL error "-10" during "" access: "SQL0010N The string constant beginning with ""XXXXT WITH CS -- OPTLEVEL( 5 ) -- QUERY_DEGREE" does not have an
ending string delimiter. SQLSTATE=42603"
/SPS_Upgrade/QS0/SUM/abap/exe/R3trans finished (0012).
INFO: No client specified on command line. Taking client 0 from E070C.
tp finished with return code: 12
meaning:
A tool used by tp aborted
SAPup> Process with PID 10823 terminated with status 12 at 20160509163851!
If we look in CP2SHDR.ELG it repeats the following:
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
COPY TO SHADOW and RETURN CODE in SAPEBBO731.QS0
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4 ETW000 [ dbntab ,00000] ***LOG BY0=>SQL0010N The string constant beginning with ""XXXXT WITH CS -- OPTLEVEL( 5 ) -- QUERY_DEGREE" does not have an
ending string delimiter. SQLSTATE=42603
4 ETW000 11 0.074003
4 ETW000 [ dev trc,00000] NTAB: db_ntab(): NT_RDTDESCR: tabname: E070~ , fieldname: , fieldnumber: 0
4 ETW000 9 0.074012
4 ETW000 [ dev trc,00000] NTAB: db_ntab(): procure_2 failed, returning 1 5 0.074017
2EETW000 sap_dext called with msgnr "1":
2EETW000 ---- db call info ----
2EETW000 function: db_ntab (shadow)
2EETW000 fcode: NT_RDTDESCR
2EETW000 tabname: E070~
2EETW000 len (char): 5
2EETW000 key: E070~
2EETW000 retcode: 1
2EETW000 sql error: -10 SQL0010N The string constant beginning with ""XXXXT WITH CS -- OPTLEVEL( 5 ) -- QUERY_DEGREE" does not have an ending string deli
miter. SQLSTATE=42603
2EETW125 SQL error "-10" during "" access: "SQL0010N The string constant beginning with ""XXXXT WITH CS -- OPTLEVEL( 5 ) -- QUERY_DEGREE" does not have an
ending string delimiter. SQLSTATE=42603"
2EETP200 Export with errors, request will not be imported
Any ideas?
Hi,
Can you restart and check it once.
Did you import anything in pre-processing phase during development lock ?
Regards,
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Having dug through the logs, we can see there's a problem surrounding DDNTT~ :
SQL0204N "SAPQS0.DDNTT~" is an undefined
SELECT COMPCNT, UNICODELG FROM "DDNTT~" WHERE TABNAME = 'SVERS' WITH UR
From other posts, I see that the above SQL statement is normally ran without the tilda on the end.
Doing a quick search for tables like DDNTT% shows DDNTT and DDNTT~~OLD - there's no DDNTT~.
But it seems that perhaps DDNTT~ is a valid table - albeit a shadow table. Looking in table PUTTB_SHD on a system we have already successfully patched, I can see that the 'SHDNAME' for DDNTT is indeed DDNTT~.
So that got me thinking... perhaps DDNTT~ does exist - but as it's part of the shadow system, does it have a different schema? Which is why we didn't see it... does the shadow instance have it's own schema, and if so, any idea of the naming convention?
Checked and DDNTT~ does NOT exist, for ANY schema.
So either
1) It's supposed to exist and has somehow got deleted
2) It's supposed to be just DDNTT without the ~ 'tilda' on the end
But which one... and if it's number 2 then where is it getting this from? The piece list transports? The ones we can see in the other systems don't contain table DDNTT
Hi Ross,
As you have mentioned in the post your SUM process got crashed. Did you check sum guide in appendix paragraph "Preparing the Restart of the Update After a System Failure"
Guide suggest some manual steps steps to be taken care in case of abnormal crash of sum process.
Please take a backup of database before you proceed and pls keep updating.
Regards,
Rajeev
Thanks - it seems that have carried out these steps.
What we then realised was that the shadow instance was down - and we are unable to start.
The error we get is:
C *** ERROR in prepareStatement[dbdb6.c, 15581] CON = 0 (BEGIN)
C &+ DbSlReadDB6( SQLExtendedPrepare 😞 [IBM][CLI Driver][DB2/LINUXX8664] SQL0204N "SAPQS0SHD.SVERS" is an undefined n
C *** ERROR in prepareStatement[dbdb6.c, 15581] (END)
B ***LOG BZA=> table SVERS does not exist on database [dblink 1327]
We checked and it seems there is NOTHING for schema SAPQS0SHD! It is completely empty....
So perhaps it reached the stage where it had finished with copying from the shadow (we see logs that suggest many tables were copied - although it did have the errors with DDNTT~)... does the shadow need to be able to start now? Or should it be finished? I'm thinking, is there a way to set a flag so SUM goes onto the next step.... although it may be it didn't fully finish and we will get problems.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Thanks but already seen that post (and a few similar ones) and it's not quite the same issue.
We have already done this work in 5 systems in 3 of the tiers and 4 in the next, so this is the 20th time we're doing this and we never had this problem before. I've checked and the same version of SUM (SUM10 SP14), R3trans and tp has been used for all systems. We tried running R3trans -x to get some more information but this actually runs successfully and ends with return code 0.
Perhaps I could look at the script that is called to see the exact SQL it's having a problem with, but I'm not sure where the find that script?
User | Count |
---|---|
87 | |
10 | |
10 | |
9 | |
7 | |
6 | |
6 | |
5 | |
4 | |
3 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.