An ODG system has been upgraded recently to 19C and redo apply stopped working. I was asked to take a look at it. It is managed by ODG broker.
When checking it in dgmgrl using “show configuration;“, I saw the error “ORA-16810: multiple errors or warnings detected for the database” on the standby database.
Then I checked the standby database with “show database “STD” ” and the actual error was “ORA-16854: apply lag could not be determined“. This message means either apply services were not running or there was no connectivity between the redo source and standby database. I tried to disable/enable the configuration and it didn’t help, still had the same error.
Then I started to check the connectivity using “dgmgrl sys/<sys_passwd>@PRI” and “dgmgrl sys/<sys_passwd>@STD” on both primary and standby servers. On the standby server, when connecting to the primary, it gave me an invalid logon error which seemed to be password for sys wrong. However, on the primary server, same password works when connecting to the primary or the standby.
So nothing wrong with the password of sys, but something wrong with connection from the standby to the primary. I looked at the tnsnames.ora on the standby, the hostname of the primary is correct, but the IP was resolved to a different IP than the actual IP of the primary. And it happened there was also an Oracle database behind that the wrong IP. That’s why I got the logon error!
It turned out DNS entry was messed up somehow and a quick fix was adding a good entry in host file on the standby server. Then got the network admin to fix the DNS entry. After that, might need to disable/enable the configuration to kick in the redo apply service or re-create the configuration if necessary.