Test Instance With Oracle 11gR2 - EM Sending Email Notifications Too Late

On a test instance with Oracle 11gR2 with Enterprise Manager Database Control I try to activate the Email Notifications in Enterprise Manager, so that I can use them for monitoring.

So i configured following settings:

Under Setup -> Notification method I gave him a valid mail address with a vaild SMTP server. The test mail was send successfully.

Under Preferences -> General i gave SYS a valid mail address to receive. The execution plan i let in standard configuration, so every time at everyday i will get an email, when something happens. the test mail was send successfully too.

Now i created an own rule to test the configuration:

I went to Preferences -> Rules an add an rule. Name and description I named "test" and set "database instance". At availability I activated all checkboxed, so i should be informed when the instance goes down, comes up, getting an error, when the error is fixed, when agent is not available, when agent is available again, when there is a metric error, when the emetic error is fixed, when a blackout starts and when it ends.

Metrics, policies and jobs there wasn't defined anything in standard, so i let it so and set nothing here. I activate the checkbox for sending emails to sys and sending repeat emails every 15 minutes 3 times at all.

Now, to test this notification i logged into my testserver and killed the listener process, to see, what will happened. After 1 Hour waiting i didn't get any email, so i restarted the listener. after the listener was started my outlook was fluted by emails from the server, that the listener was down. But why i didn't get them earlier.

I checked the procedure with the database to, here i get it down by "shutdown abort" and "shutdown immediate". But i get the mails, that the database is down, when its up already. And thats to late!

When I want to use such email notifications for productive servers and someting is going wrong, for that a notifications is configured of course, I want to get the mail directly, when that happened and not, hen the problem is fixed!

For example, here an email from testing, where the listener was down, that i've get at 13:01 :

Target Name=LISTENER_<hostname>
Target Type=Listener
Host=<hostname>
Metric=Status
[code].........