Hi Mario,
At my customer side we are in a DEVOPS approach and we are using JIRA.
They are just tools and approaches made to accelerate and to shorten the availability of functionalities or treatments. SCRUM is the same notion (a kind of a former word for the same queries to shorten time to market or time to user access ...).
JIRA can be used to specify for DEV, for OPS, to supply details of needs from FUNC(TIONAL) to other FUNC (subcontractors) but not to describe with a clean why a well-composed requirement.
Thus, devops/scrump people are kind of people just coming from the develpment world / infra production (ITIL way) but the willing to develop small components will create a kind of a forest that they want to control by API MGT tools ... for instance
Then , more and more Architecture is important / mandatory if not ... :
- a true requirement management (not JIRA that is a simple need repository, but modelized requirements what you can do with MEGA-HOPEX and with your requierments linked to all layers of the architecture ...)
- a true data modelisation to share it because with SCRUM / DEVOPS data are exposed to be exploded everywhere.
- a functional architecture to be matured to know "what" for "who" - "with what" ?
etc.
There are then a great risk / peril.
DEVOPS / SCRUM must be absolutly architectured then ...
What is nice is to link HOPEX requirements with JIRA items ... for instance.
We want to put HOPEX requirements id in RANOREX scenarios too ... (http://www.ranorex.com/)
MEGA / SCRUM : to change the right models to modelize it rapidly, have an access to master data (data models, list of functionalities, BPMN, ...). Import the API MGT list of services to parse it and have a functional dictionnary of API inside HOPEX etc.
An infinite universe indeed ... for nearly no overlap between devops tools like JIRA and modelization tools ... like HOPEX
Lionel Mazurié
