16. models@run.time : body of work
IEEE Computer: Special issue on
Models@run.time, 2009
Computing: Special issue on Models@run.time,
2013
Workshop models@run.time at MODELS (12th
edition since 2006)
http://st.inf.tu-dresden.de/MRT17/?site=editions
Workshop models@run.time at ICAC (2nd edition)
Dagstuhl Seminar – 2011
Book Models@run.time - Foundations, Applications,
and Roadmaps - 2014
21. Models@run.time
to Support Synthesis of Emergent Middleware
[Compu3ng 2013]: The Role of Models@run.1me in Suppor1ng Synthesis of Emergent
Middleware, Bencomo, Bennaceur, Grace, Blair, Issarny
EU ConnectProject
22. Models@run.time
to Support Synthesis of Emergent Middleware
[Compu3ng 2013]: The Role of Models@run.1me in Suppor1ng Synthesis of Emergent
Middleware, Bencomo, Bennaceur, Grace, Blair, Issarny
EU ConnectProject
27. • The system has adapted to a new
configura1on
– what? how? why?
• The systems has crashed when trying
to adapt
– what? how? why?
• The user may want to understand the
system in terms of requirements or
their own language
Very (sad) user
A well-known problem with self-adap1ve autonomous
systems is that users may not understand or trust them
28. • The system has adapted to a new
configura1on
– what? how? why?
• The systems has crashed when trying
to adapt
– what? how? why?
• The user may want to understand the
system in terms of requirements or
their own language
Very (sad) user
A well-known problem with self-adap1ve autonomous
systems is that users may not understand or trust them
The system needs to offer explana1ons
using the language of the stake holder
40. U
Evidence
Collect Data
Frequently (D)
Energy
Efficiency (E)
Decision
SP FH
22
Dynamic Decision Networks frame
the decision-making of a self-adap1ng system
Decisions (goal realizations)
SP: Clean when Empty SH: Clean at Night
Chance node) (Softgoals - non functional requirements)
M : Minimize Energy Cost A : Avoid Tripping Hazard
P(D|SP)
[REFSQ 2013] – Best Paper Award Bencomo et all
[SEAMS 2013] Bencomo, Belaggoun, Issarny
[SEAMS 2015] Hassan, Bencomo, Bahsoon
EUj
= EU(dj
| e) = P(xi
i
∑ | e,dj
)U(xi
| dj
)
j = 1,2...
• Types of nodes:
• Chance nodes: labeled by random variables Xi that
represent the states of the world (NFRs)
• Decision nodes: with the set of configura1ons
• U;lity nodes: that state the preferences
about the states of the world
• Evidence nodes: to denote the observable variables
The condi3onal probabili3es quan3fy the effects of
decisions on states of the world
41. X1(t) X(t+1)
D(t) D(t+1)
U(t+1)U(t)
E(t) E(t+1)
Evidence
depends
on state
X2
X2
….
….
….
Time t Time t+1 Time t+n
Dynamics Decision Networks (DDNs)
EUj
= EU(dj
| e) = P(xi
i
∑ | e,dj
)U(xi
| dj
)
j = 1,2...
47. Remote Data Mirroring (1)
Copies of important data are stored at one or more secondary loca1ons
Goal: Protect data against loss and
unavailability
Case Study
• Design choices
• Remote mirroring protocols
e.g. Minimum spanning tree (MST) vs Redundant topology (RT)
(1) “Relaxing claims:Coping with uncertainty while evalua3ng assump3ons at run 3me,” A. Ramirez, B. Cheng, N. Bencomo,
and P. Sawyer, ACM/IEEE Int. Conference on Model Driven Engineering Languages & Systems MODELS, 2012.