Design, Implementation and Evaluation of a Storage System for Delay-Tolerant Networks

Simple applications of networked information technology have been shown to have an impact in the developing regions in the areas of health care, education, commerce and productivity. However, use of information technology in developing regions setting is hampered by a lack of inexpensive reliable telecommunications infrastructure. In particular, existing applications relevant to this space are built using software architectures which assume always-on, low-latency, end-to-end connectivity. One potential connectivity solution is the Delay-Tolerant Networking (DTN) stack, an overlay network protocol that can route across network partitions and heterogeneous forms of network infrastructure. However, the DTN message-based interface is a poor fit for many applications as they are more naturally structured in terms of shared state. To address these issues, we present TierSync, a distributed eventually-consistent shared-storage synchronization primitive for DTNs. TierSync enables applications to share persistent data among TierSync nodes in an efficient and flexible manner. Novel features of the TierSync protocol include efficient support for fine grained partial sharing and the ability to arbitrarily order updates for data prioritization. We demonstrate an implementation of the TierSync protocol as a file-system and show useful applications can be easily ported to the TierSync system.

[1]  Umesh Maheshwari,et al.  Garbage collection in a large, distributed object store , 1997 .

[2]  Rabin K. Patra,et al.  Using redundancy to cope with failures in a delay tolerant network , 2005, SIGCOMM '05.

[3]  Douglas B. Terry,et al.  Peer-to-Peer Replication in WinFS , 2006 .

[4]  Michael J. Demmer,et al.  DTLSR: delay tolerant routing for developing regions , 2007, NSDR '07.

[5]  Arun Venkataramani,et al.  DTN routing as a resource allocation problem , 2007, SIGCOMM '07.

[6]  Marvin Theimer,et al.  Managing update conflicts in Bayou, a weakly connected replicated storage system , 1995, SOSP.

[7]  Arun Venkataramani,et al.  DTN routing as a resource allocation problem , 2007, SIGCOMM 2007.

[8]  Ashish Goel,et al.  Perspectives on optimistically replicated, peer‐to‐peer filing , 1998, Softw. Pract. Exp..

[9]  John S. Heidemann,et al.  Resolving File Conflicts in the Ficus File System , 1994, USENIX Summer.

[10]  Peter L. Reiher,et al.  Roam: A Scalable Replication System for Mobility , 2004, Mob. Networks Appl..

[11]  Jörg Ott,et al.  Integrating DTN and MANET routing , 2006, CHANTS '06.

[12]  Jung A Kim Telehealth in the Developing World , 2010, Healthcare Informatics Research.

[13]  Paul Mackerras,et al.  The rsync algorithm , 1996 .

[14]  Eric A. Brewer,et al.  Harvest, yield, and scalable tolerant systems , 1999, Proceedings of the Seventh Workshop on Hot Topics in Operating Systems.

[15]  Dennis Shasha,et al.  The dangers of replication and a solution , 1996, SIGMOD '96.

[16]  Marvin Theimer,et al.  Flexible update propagation for weakly consistent replication , 1997, SOSP.

[17]  Lakshminarayanan Subramanian,et al.  Rethinking Wireless in the Developing World , 2006, HotNets.

[18]  Steven McCanne,et al.  Scalable data naming for application level framing in reliable multicast , 1998, MULTIMEDIA '98.

[19]  Eric Brewer,et al.  Implementing Delay Tolerant Networking , 2004 .

[20]  Lei Gao,et al.  PRACTI Replication , 2006, NSDI.

[21]  Dahlia Malkhi,et al.  P2P replica synchronization with vector sets , 2007, OPSR.

[22]  Eric A. Brewer,et al.  Analysis of WWW traffic in Cambodia and Ghana , 2006, WWW '06.

[23]  Kendall Ho,et al.  Telehealth in the Developing World , 2009 .

[24]  Bo Leuf,et al.  The Wiki Way: Quick Collaboration on the Web , 2001 .

[25]  M. Frans Kaashoek,et al.  Rover: a toolkit for mobile information access , 1995, SOSP.

[26]  Mahadev Satyanarayanan,et al.  Disconnected Operation in the Coda File System , 1999, Mobidata.

[27]  Peter L. Reiher,et al.  Rumor: Mobile Data Access Through Optimistic Peer-to-Peer Replication , 1998, ER Workshops.

[28]  Robert Wilensky,et al.  The hash history approach for reconciling mutual inconsistency , 2003, 23rd International Conference on Distributed Computing Systems, 2003. Proceedings..

[29]  Mahadev Satyanarayanan,et al.  Flexible and Safe Resolution of File Conflicts , 1995, USENIX.

[30]  Mike Hibler,et al.  An integrated experimental environment for distributed systems and networks , 2002, OPSR.

[31]  Kevin R. Fall,et al.  A delay-tolerant network architecture for challenged internets , 2003, SIGCOMM '03.

[32]  Akihiro Nakao,et al.  Postmanet: turning the postal system into a generic digital communication mechanism , 2004, SIGCOMM 2004.

[33]  Eyal de Lara,et al.  Haggle: Clean-slate networking for mobile devices , 2007 .

[34]  R. Durst,et al.  Interplanetary internet (ipn): architectural definition , 2001 .

[35]  Ben Y. Zhao,et al.  Pond: The OceanStore Prototype , 2003, FAST.

[36]  Sanjeev Khanna,et al.  Agreeing to Agree: Conflict Resolution for Optimistically Replicated Data , 2006, DISC.

[37]  Irene Greif,et al.  Replicated document management in a group communication system , 1988, CSCW '88.

[38]  Benjamin C. Pierce,et al.  What's in Unison? A Formal Specification and Reference Implementation of a File Synchronizer , 2004 .

[39]  Srinivasan Keshav,et al.  Low-cost communication for rural internet kiosks using mechanical backhaul , 2006, MobiCom '06.

[40]  Dahlia Malkhi,et al.  Concise version vectors in WinFS , 2005, Distributed Computing.

[41]  Mostafa Ammar,et al.  Multicasting in delay tolerant networks: semantic models and routing algorithms , 2005, WDTN '05.

[42]  Arun Venkataramani,et al.  Interactive wifi connectivity for moving vehicles , 2008, SIGCOMM '08.

[43]  Mark Halsey,et al.  Searching the World Wide Web in Low-Connectivity Communities , 2002 .

[44]  Werner Vogels,et al.  Dynamo: amazon's highly available key-value store , 2007, SOSP.

[45]  Dirk Kutscher,et al.  Scalable DTN distribution over uni-directional links , 2007, NSDR '07.

[46]  Richard A. Golding,et al.  Weak-consistency group communication and membership , 1992 .

[47]  Butler W. Lampson,et al.  Designing a global name service , 1986, PODC '86.

[48]  Nancy A. Lynch,et al.  Brewer's conjecture and the feasibility of consistent, available, partition-tolerant web services , 2002, SIGA.

[49]  Dan Walsh,et al.  Design and implementation of the Sun network filesystem , 1985, USENIX Conference Proceedings.

[50]  Stephen Farrell,et al.  Bundle Security Protocol Specification , 2011, RFC.

[51]  Alley Stoughton,et al.  Detection of Mutual Inconsistency in Distributed Systems , 1983, IEEE Transactions on Software Engineering.

[52]  Alex Pentland,et al.  DakNet: rethinking connectivity in developing nations , 2004, Computer.

[53]  Catherine C. Marshall,et al.  Cimbiosys: a platform for content-based partial replication , 2009, NSDI 2009.

[54]  John S. Heidemann,et al.  The Ficus Replicated File System , 1992, OPSR.

[55]  Eric A. Brewer,et al.  The case for technology in developing regions , 2005, Computer.

[56]  Jennifer Vesperman Essential CVS , 2003 .

[57]  Sape Mullender,et al.  Distributed systems , 1989 .