Book a Demo

Author Topic: Time constraint on a Use Case  (Read 3512 times)

Gary

  • EA User
  • **
  • Posts: 84
  • Karma: +1/-0
    • View Profile
Time constraint on a Use Case
« on: November 24, 2014, 10:41:33 pm »
What is the best way to show that a use case has an execution time limit?
I know I can show this on the sequence diagram that elaborates a use case. However this will not show up on a relationship matix.

So I have 2 requirements:
The system shall perform x functionality.
X functionality shall complete in less that 3 seconds.

I create a UseCase "Perform X" and link with a <<refine>> (SysML)
Now what is the best way to link the constraint requirement to the usecase? Or should I even try?
I can create a testcase and state that a condition of succes is that it takes less than 3 seconds. Again though there is no explicit link between the requirement and the usecase.
The reason I ask is that I may have to hand over the implementation to someone else and would like to include all of the requirememnts that relate to the use case. Plus it seemed like an interesting question?

Any suggestions?

Gary

philchudley

  • EA User
  • **
  • Posts: 750
  • Karma: +22/-0
  • EA Consultant / Trainer - Sparx Europe
    • View Profile
Re: Time constraint on a Use Case
« Reply #1 on: November 24, 2014, 10:58:14 pm »
One possibility is to use a SysML Requirement of the type

[highlight]<<designConstraint>>[/highlight]

which is supported in EA

In the description enter your timing constraint(s)

Using the relationship matrix it is a simple matter to create SysML [highlight]satisfy[/highlight] relationship(s) between all such <<designConstraints>> and corresponding Use Cases

Cheers

Phil
Models are great!
Correct models are even greater!

Gary

  • EA User
  • **
  • Posts: 84
  • Karma: +1/-0
    • View Profile
Re: Time constraint on a Use Case
« Reply #2 on: November 25, 2014, 12:39:36 am »
Phil

Yes I did think about the <<satisfy>> link. I know that the client end can be any element as stated in the SysML spec, but I always fell a bit twitchy about using it for any thing other than block to requirement linkage. I did also consider adding a statement into the Usecase constraints as a process type.

anyone else have any thoughts?

Gary