Recently we have patched our SAP ECC system EHp 4.0 to EHp 5.0. EHP 5.0 requires Netweaver 7.0 EHP 2.0. As this is not a release upgrade but it still requires lot of efforts & time to complete, that’s why it’s call upgrade. We have calculated our STACK.xml file from SOLMAN 7.0 EHP 1 SP 23 (EHP 5.0 requires minimal level SP23). we stuck in lot of errors and it takes time to resolve it.
Before starting EHP upgrade always prepare a download directory and keep your all downloaded SAR & stack.xml files in same directory. During upgrade sometimes EHPi got hangs & you feel it’s not working so please check abap/log or abap/tmp for any log or errors.
Below is the quick glance of errors which i faced during upgrade.
Go through this below SAP note before starting EHP upgrade.
|1302772||General Installation Note|
1. Extraction Phase – Missing Notes – It will show you some missing notes. you have to import SAP notes in your ECC 6.0 EHP 4.0 system via SNOTE to continue. (some notes can only be applied in client 000 & can also requires functional people so make sure during note implementation you get confirmation from functional team)
2. Configuration Phase – SPAM Version Check - To avoid this error update your SPAM to latest before starting upgrade.
3. Configuration Phase – Include Change Request – if you want to include any change request to be imported during upgrade to give details otherwise continue to skip this option.
4. Configuration Phase – Errors - Download latest R3trans, tp & libdbsl(dbsoralib.dll) from Marketplace and extracts it to EHPI/abap/exenew & EHPI/abap/exe dir. This is the place where your new/extracted kernel resides, recheck that you are downloading the right R3trans & tp for your new kernel & also check R3trans -d( it should finish with 000) from your new kernel dir( i. e. abap/exe or abap/exenew)
5. Check Phase – Errors – Create Tablespace PSAPSR3702 with <given space>. This is the manual activity. you have to complete this activity via BRtools. You also get one file with name “ORATBSXT.LST“ in abap/log, in this file you will find how to create tablespace with recommended size.
6. Pre processing Phase - Open repairs found – If your system is development system or any unreleased requests are there so release it or before proceed.
7. Pre processing Phase – MAIN_INIT/JOB_RSVBCHCK2 were Negative- Check log file and find the errors. this error occurs if you have any update request pending in SM13 or any outbound queue pending in SMQ1. Delete all the request from SM13 & SMQ1 and proceed. if it’s still giving error also run report RMCEX_SETUP_ENTRIES and delete the structural data or call transaction LBWG to delete the contents of the setup table for all clients. Also check this SCN post http://scn.sap.com/thread/903136
8. Pre processing Phase – Lock development – This option will lock your development system for future development till upgrade.
9. Pre processing Phase – MAIN_SHADOW/START_SHDI_FIRST – failed to start shadow instance, check log file STARTFSI.log & DEVTRACE.log. usually this error occurs where there is short memory. in our case we check that if we down our main instance then shadow instance get up. both are not working simultaneously. we insert the parameter in shadow instance profile. (abap/heap_area_total = 2000000000) and also set the recommend page file size and restart the shadow instance. also check this SCN post – http://scn.sap.com/thread/1876639
10. Pre processing Phase – ACT_UPG – in our case ACT_UPG phase takes more then 12 hours to complete.
11. Take full backup before downtime starts.
12. Downtime Phase – MAIN_SWITCH/KX_SWITCH_1 - check log files, in our case EHPi fails to copy to 2 kernel files from exenew dir to /usr/sap/<SID>/…. we manually copy that 2 files and proceed.
13. Downtime Phase – MAIN_NEWBAS/STARTSAP_NBAS! - check log files, in our case SAP error reading our instance profile from /usr/sap/<SID>/sys/profile. There was junk character in the profile but after removing that character EHPi again not able to read the profile so we copy the contents of our instance profile and create a new instance profile and then proceed.
14. No more Errors in Downtime & Post Processing Phase.
We have completed our ECC 6.0 EHp 4.0 to EHp 5.0 successfully after these efforts. i was not able to note down the exact time for every phase.
Manual Post steps – in our case after logging to EHp 5.0 SAP system, we found that our Updates are deactivated, we activate the updates in SM13 & also create structure of financial. ( i forgot the exact thing as this is done by my colleague.)