Self resolved alert after some time.

[cloud-digitaltag]

Hi.
I have a wrote alert for process running:
(count by (process_name,instance, environment) (check_cpu_process) or on() vector(0)) < 1

When I stoped a process for tests, I see that alert firing, and everything ok. But after for example 9 minutes I see that alert is resolved.
On Grafana explorer when I run above expression, I have data:

but when I look at the alert (view rule) I have no data:

and I'm wondering, what I did wrong.
Unfortunately I have only this metrics, and I'm trying from cpu data create alert, that some process not running.
my idea is that I count scrapes and when is no data for 1m I firing them. Maybe is another way to prepare alert. I tried also absent but not working, because I have the same process name on different instances and absent get me no data.

Thanks in advance for help.