Rudder

When we use the "catchup" mode in schedule_* , it will run the job asap if the job was never run. And then define a persistent class for the length of the desired frequency
So when it will be the desired period for this job, the class will still be set, and the job won't be run. After the end of the persistent class, the job will be caught up

The result is a job not splayed at all

Impact is worsen by the fact that it is used by rudder inventory splay mechanism