Oracle Database 11g Enterprise Edition Release 18.104.22.168.0 - 64bit Production, Running on Oracle Linux 6.5, we configured full RMAN backup schedule using OEM (
Oracle Enterprise Manager 11g Database Control Release 22.214.171.124.0) to take full backup once a week, and incremental level 1 for the rest of the week. This setup went smoothly until we were forced by internal policy to change passwords.
We did change the host's credentials (
oracle user and
root) plus the main schema.
system were unchanged (yet), then I did update the preferred credentials in the OEM, but the next job failed!
Before retrying, I double checked that we chose to use the preferred passwords, and that I didn't misspelled when I typed them (I did
test the preferred credentials, misspelled host, sysdba, and normal credentials and watched the errors for each, then entered them carefully and tested again and had a success), but the job failed again. I did change the credentials for the job so it use custom credentials and entered the new credentials manually, retried and failed again!
To get more sure, I created a new job, the same settings we chose for the old job, using the preferred credentials, and ran it immediately, and it succeeded! This made me sure that I didn't mess up the preferred credentials, but there's something in the old job that doesn't respond the way I expect it to do, which is to take the preferred credentials at runtime. It even didn't take the custom credentials when I manually specified it!
So what am I missing here? Is there any other step that I overlooked or forgot to do (or didn't know I should)? Is it the expected behavior and I should delete and recreate all the jobs whenever I change passwords?
OEM version (output of
Enterprise Manager 11g DB Control Version 126.96.36.199.0 Enterprise Manager 10g Agent Version 10.2.0.4.0