We intensively prepare the upgrade of our Multitenant 12.1 CDBs to Oracle Release 12.2.
I have already blogged about the big advantage of the new version compared to the current 12.1
Oracle Multitenant 12.2 – time to rethink the schema based consolidation?
It’s clear that we want to migrate our current containers to the new version as fast as possible.
To register 12.2 databases in the recovery catalog you have to upgrade the catalog. But be careful: up to this day it contains a bug.
There’s quite a high probability that you can’t sync your 12.1 CDBs with the RMAN Catalog after the upgrade!!!
RMAN> resync database; starting full resync of recovery catalog RMAN-00571: =========================================================== RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS =============== RMAN-00571: =========================================================== RMAN-03002: failure of register command at 04/25/2017 14:55:13 RMAN-03014: implicit resync of recovery catalog failed RMAN-03009: failure of full resync command on default channel at 04/25/2017 14:55:13 ORA-20110: set stamp set count conflict RMAN> exit
I have an open SR and the following patch should fix the error.
19209117: Tracking int dif in SRG7: RMAN-3008: Error while performing automatic resync
I’m waiting for Oracle support to send me the version that matches our patch level. I will update the blog entry as soon as I know if the patch works or not.
In the meantime, you might rather create a separate catalog for the 12.2 databases as a temporary workaround.
Update on 22 June 2017 ————————————————————————————————–
So Oracle provided to install the patch
At least in my enviroment it still not working to sync the 12.1 cdb with the 12.2 catalog 🙂
So the SR has now severity 1 on oracle support, and is monitored by the Vice President for Database Upgrade and Utilities. So i hope that i get the final solution in the next days.
Update on 31 July 2017 ————————————————————————————————–
So finally oracle found the solution for the problem with the “ORA-20110: set stamp set count conflict”. I’ve implemented it three days ago and over the weekend all backups rund without error and the catalog is up to date 🙂
Here is the statement from oracle support
Hi, Problem here is the in-correct join in cursor bs where we missed to join the 2 tables ie v$backup_set & v$containers . <<recover.bsq_12102_160419_OK>> where bs.recid between low_recid and high_recid and (bs.stamp >= kccdivts OR bs.recid = high_recid) and bs.stamp >= resyncstamp and bs.for_xtts != 'YES' and bs.con_id = pdb.con_id(+) << order by bs.recid; <<recover.bsq_12102_161018_NOT_OK>> from v$backup_set bs, v$containers pdb where bs.recid between low_recid and high_recid and (bs.stamp >= kccdivts OR bs.recid = high_recid) and bs.stamp >= resyncstamp and bs.for_xtts != 'YES' order by bs.recid;
So in my case the recover.bsq file under $ORACLE_HOME/rdbms/admin/ had the wrong join operation. So oracle send us a modified recover.bsq file and afterwards everything works fine.
So if you ran into the same error open an SR on oracle Support and tell them, that you need a fixed version of recover.bsq file that matches on your PSU or Bundle Patch level