Dg11G: Primary database loses dest_10 parameter when becoming primary
Hello everyone
I have a little bit of strange situtation here. I have recently upgraded a 10g database to 11.2.0.3 and also reconfigured data guard to match out new standard.
Everything is working just fine, except one odd behavior. When I perform a switchover, the dgbroker (I guess) resets parameter log_archive_dest_10 for me to '' , which means my database will not be able to write any archivelog and will end up with archiver error.
The setup is 3 nodes, currently in maximum performance but will be a maximum protection once it's stable with the switchovers. Have searched half the internet and scratched my head for hours trying to solve this.
At first there was this issue on all nodes , but after cleaning up the spfiles a bit I managed to keep the dest_10 on 2 of the nodes, but when I switch to the 3rd, the issue comes back.
Everything else seems to work fine, redo logs are shipped and applied and switchover works fine.
My log_archived_dest_10 looks like this:
log_archive_dest_10='location="USE_DB_RECOVERY_FILE_DEST", valid_for=(ALL_LOGFILES,ALL_ROLES) db_unique_name="DBNAME_host1"' scope=spfile;
(The database is not actually named DBNAME nor host1 , but to make sure I don't violate any policy I changed the name)
Maybe I understood something wrong when setting all this up, but we have a bunch of DG's that is using the same type of setup and they work fine. Only thing that is different is that this was a Oracle DG 10G before we upgraded.
There is no observer connected.
Hope anyone can help out and thanks in advance.
Sincerely
Johnny