| |
Comment: |
This looks more like Requirements than Use Cases, but it is all right, there is really USEFUL TECHNICAL INFORMATION inside.
It is absolutely needed that you describe :
- the SEQUENCE OF INTERACTIONS between the ACTORS and the SYSTEM.
- WHAT is the system, and which are its BOUNDARIES (INTERFACES with the actors). In particular, does the system expose each local storage of each computing resource as interface ?
The execution service can easily return the state (and other information) of a job queried by its ID. But querying jobs by owner, date, or complex selection really has impact on performances, and should probably be handled by a separate service.
The execution service already has a lot of work managing jobs, which are transient entities. Managing job logs as persistent data, and responding to complex queries, is a completely different business, and should probably be handled by a different service.
|
