Please check the current total active job count.
In particular, check the cumulative number of failed jobs.
In my case, there was the same problem, and there were about 2,000 failed jobs, but after clearing this job, this phenomenon disappeared.
I recommend that you delete the failed job. (Change failed job to completed).
In addition, in situations where there are a lot of failed jobs, Oracle redolog occurs tremendously.
I suffered a lot because I did not know the cause of this problem (a huge redo log occurred).
I don't know why the DE architecture is like that.