Rich, have you tried a somewhat brute-force method, checking all config and
properties files involved to see if any reference the old server? Our 12c
EMs have a root directory under which gc_inst, MW and swlib reside.
Searching at this point I find the following "*.config" and ".properties"
files with my EM server name in:
./gc_inst/em/EMGC_OMS1/emgc.properties
./gc_inst/em/EMGC_OMS1/sysman/ocm/emoms.properties
./gc_inst/NodeManager/emnodemanager/nodemanager.properties
./gc_inst/user_projects/domains/GCDomain/servers/EMGC_OMS1/data/nodemanager/startup.properties
./gc_inst/WebTierIH1/config/OPMN/opmn/instance.properties
./MW/oms/ocm/repeater/config/ocm_config.properties
./MW/utils/ccr/domainlocation.properties
Dave
On Mon, Oct 2, 2017 at 9:30 AM, Rich J <rjoralist3@xxxxxxxxxxxxxxxxxxxxx>
wrote:
Hey all,
I'm working on upgrading EM 12.1.0.5 to 13.2 by migrating the OMS and
repository DB from a single OL5 server to a single OL7 server.
The first step for me was to upgrade the repo DB from 11.2 to 12.1, and
then migrate it to the OL7 box. I used the instructions at
http://docs.oracle.com/cd/E24628_01/install.121/e24089/
ha_backup_recover.htm#EMADV10748 ("19.6.2.3 Full Recovery on a Different
Host") for the migration. Seeing as the instructions failed to mention the
need to backup and restore the emkey, I'm not sure it's complete.
While everything's generally working after the repo DB migration, I
noticed that the repository DB *target* still thinks it's on the original
server. And the Agent on the original server also thinks the repo DB is
still there, even though I had "deinstalled" that $OH. EM's running and
collecting metric data and alerting like usual, and it can even monitor the
repository as a target, but some pieces still don't know that it's been
migrated to a new server.
I wanted to get this fixed before attempting a migration/upgrade of the
OMS, but after a 5 week SR, Support says the fix is to migrate the OMS.
Thoughts anyone?
Thanks,
Rich
