question

If you need a setup event totally independent by the current event of the processor, you can use time tables. For each different down time you add another time table.

If you are familiar with process flow, you add a Date Time source activity. After that you create for each processor a token by a create token activity or you change the amount of created tokens in the source activity. Please look into the source code editor to add labels in the source or create tokens activity just to identify the source code header and the declaration of default variables. There is certainly a variable which enumerates a value of current created token like “creationRank”. This variable name is set as a value for a label at the token. I remind you later why you do this.
Then there is a sequence of a custom code, a delay and another custom code activity. The first custom code stops an object, the second let’s resume the same object. Stop and resume are picklist templates of a custom code activity. The reference of the involved object is a label value at the token. The delay time is also a label value at the token. You setup those label values by lookup of a global table where you refer to the involved object and delay time. The label of the “creationRank” is a reference to the row of the global table. Columns are involved object and delay time.