Date

March 5-6th, 2012

Where



LORIA
Centre de Recherche INRIA Nancy-Grand Est
615, rue du Jardin Botanique
54600 Villers-lès-Nancy

Meeting will take place in Room B013-B011

LORIA / INRIA Nancy - Grand Est

Time Schedule

Monday, March 5th

  • Meeting from 9:30am to 6:00pm
  • Dinner at 8:30pm

Tuesday, March 6th

  • Meeting from 9:30am to 5:00pm

Participants

SCORE: Claudia-Lavinia Ignat, Gérald Oster, Hien Thi Thu Truong, Luc André, Hyun-Gul Roh, Pascal Urso, Stéphane Martin, Mehdi Ahmed-Nacer
REGAL: Marc Shapiro, Ikram Chabbouh, Annette Bieniusa
CASSIS: Asma Berregba
ASAP: Stéphane Weiss
XWiki SAS: Fabio Mancinelli
GDD:

Meeting Minutes

Feel free to take/read notes during the meeting using the pad available at the following url:

http://sourceplant.loria.fr/p/streams-demo (login: streams / password:minutes)

Minutes

Sessions (preliminary program)

Sessions will be dedicated to scientific and technical presentations. Other slots will be dedicated to discussions on progress, deliverables, next actions, etc.

Day One

9:30am 10:00am Handshaking Session
10:00am 10:30am Opening Session
10:30am 12:30pm Task 2: Requirements and Infrastructure
12:30pm 1:30pm lunch
1:30pm 3:30pm Task 3: Efficient Replication and Consistency Maintenance
3:30pm 4:00pm break
4:00pm 6:00pm Task 4: Security and Privacy

Day Two

9:30am 11:00am Task 5: Experimentations and Performance Analysis
10:30am 11:00am break
11:00am 12:30pm Discussions
12:30pm 1:30pm lunch
1:30pm 3:30pm Discussions
3:30pm 4:00pm break
4:00pm 5:00pm Closing Session


Day One

Task 2: Requirements and Infrastructure

JMPBCast: A New Gossip-based Dissemination Protocol for Ubiquitous Collaboration.

Presenter: Hyun Gul Roh Documents: PDF

Abstract. As novel types of interactive and mobile devices, such tablets, interactive whiteboards, and slate PCs, are catching on, people get to have three powers, such as mobility, interactivity, and connectivity. These increasing powers obtained from those devices will lead people to ubiquitous collaboration, which means anytime anywhere collaboration with any size of group. The ubiquitous collaboration is expected to change and improve the situation of the education, business, and medical environments.

For ubiquitous collaboration, interactivity can be handled by CRDTs, which are a class of operation-based replication. When CRDTs are taken into consideration, connectivity has not been deeply discussed yet. For ubiquitous collaboration supported by CRDTs, scalable and reliable connectivity is highly required because CRDTs exchange plenty of operations among nodes. Lightweight probabilistic broadcast (LPBCast), a representative gossip-based dissemination protocol is known to be scalable and reliable. This talk will present the experience learned from the implementation of LPBCast with real-time collaboration workloads.

Based on the experience of LPBCast, I propose a new gossip-based dissemination protocol named JMPBCast. JMPBCast is more systematically designed probabilistic broadcast on the basis of LPBCast. For ubiquitous collaboration supported by operation-based replications such as CRDTs, JMPBCast is more reliable, and provides the initialization of a new member without heavy space overhead.

Architecture for Optimistic Replication over P2P networks

Presenter: Stéphane Weiss Documents: PDF

Abstract. TBA

Task 3: Efficient Replication and Consistency Maintenance

Conflict-free Replicated Sets

Presenter: Annette Bieniusa Documents: PDF

Abstract. The talk will discuss the design of a conflict-free replicated distributed sets. A set is useful in many applications, and serves as a building block for other structures such as maps or graphs. We propose a theory combining state- and operation-based propagation of updates, and a novel concurrent specification approach, strengthening the theory of CRDTs. Based on a concurrent specification of sets, we explain the flaws in previous replicated set designs, and propose provably correct alternatives.

On the Undoability Problem in Distributed Collaborative Applications

Presenter: Asma Berregba Documents: PDF

Abstract. Collaborative applications are important programs allowing many users to simultaneously cooperate in order to perform a given task on a shared object. These applications are based on replicating shared objects in order to ensure data availability. The Operational Transformation (OT) approach is used for supporting optimistic replication in these applications. It allows the users to concurrently update the shared data and exchange their updates in any order since the convergence of all replicas, i.e. the fact that all users view the same data, is ensured in all cases. Undoing operations is an indispensable feature for many collaborative applications mainly collaborative editors. It provides the ability to restore a correct state of the shared data after erroneous operations. In particular, selective undo allows users to undo any operation and is based on rearranging operations in the history. Combining OT and undo approaches is a challenging task. Even though various undo solutions have been proposed over the recent years, verifying their correctness is still a challenging problem due to the absence of formal guidelines for undo. In this presentation, we address the undo problem from a theoretical point of view. As the main contribution, we propose a necessary and sufficient condition for undoing replicated objects based on OT with respect to three inverse properties. To overcome the difficulty of necessity proof, we use Constraint Satisfaction Problems (CSP) theory in order to cover all possible transformation cases. As the main result, we prove that it is impossible to achieve a correct undo for applications based on non commutative operations.

Using Telex in Disconnected Cooperative GIS

Presenter: Ikram Chabbouh Documents: PDF

Abstract. This talk is about using Telex for disconnected cooperative engineering in the context of a Geographical Information System called Visit Anywhere. Telex is a generic platform that eases development of collaborative applications by taking care of replication, consistency, conflicts, and commitment across the distributed application servers. The presented work is still preliminary, so I'll be mostly presenting the general context, the expressed needs, the difficulties encountered so far, and the preliminary solution we are working on.

Task 4: Security and Privacy

Access Control for Distributed Collaborative Editors

Presenter: Asma Berregba Documents: PDF

Abstract. The importance of collaborative systems in real-world applications has grown significantly over the recent years. The majority of new applications are designed in a distributed fashion to meet collaborative work requirements. Although such applications are more and more used into many fields, the lack of an adequate access control concept is still limiting their full potential. In fact, controlling access in a decentralized fashion for such systems is still a challenging problem, as they need dynamic access changes and low latency access to shared documents. In this presentation, we propose a generic Multi-Administrator Access Control based on replicating the shared document and its authorization policy at the local memory of each user. We consider the propagation of authorizations and their interactions. We propose a generic approach to enforce access control in existing collaborative editing solutions. Finally, we apply our framework on a collaboration prototype and measure its performance in the distributed grid GRID5000 to highlight the scalability of our solution.

Securing Logs in Operation-based Collaborative Editing

Presenter: Claudia-Lavinia Ignat Documents: PDF

Abstract. In recent years collaborative editing systems such as wikis, GoogleDocs and version control systems became very popular. In order to improve reliability, fault-tolerance and availability shared data is replicated in these systems. User misbehaviors can make the system inconsistent or bring corrupted updates to replicated data. Solutions to secure data history of state-based replication exist, however they are hardly applied to operation-based replication. In this paper we propose an approach to secure log in operation-based optimistic replication system. authenticators based on hash values and digital signatures are generated each time a site shares or receives new updates on replicas. authenticators secure logs with security properties of integrity and authenticity. We present in detail algorithms to construct and verify authenticators and we analyse their complexities.

Day Two

Task 5: Experimentations and Performance Analysis