SBAM is the middleware directly coming from results of the ANR project SPADES. SBAM initiates a non-intrusive, but highly dynamic environment able to take advantages of available resources without disturbing their native mechanism. SBAM federates multisite resources in order to schedule, submit and compute users' tasks in a transparent way.

Please use this reference to cite SBAM :

@article{FGCS2012,
  author    = {Eddy Caron and
               Florent Chuffart and 
               Cedric Tedeschi},
  title     = {{When Self-Stabilization Meets Real Platforms: an Experimental Study of a Peer-to-Peer Service Discovery System}},
  journal   = {Future Generation Comp. Syst.},
  url       = {http://dx.doi.org/10.1016/j.future.2012.10.003}
  year      = {2012},
  pages     = {a paraitre},
}

Overview

SBAM is, firstly, a decentralized grid middleware. It relies on a P2P approach, i.e., a set of agents able to discover resources and schedule computing tasks over a federation of heterogeneous computing platforms (petascale computers, data centers, clouds…). SBAM dynamically acquires and releases resources of computing sites according to users' needs and conditions, to federate them into a global constantly growing or shrinking logical platform, referred to as the overlay.

SPADES

The SPADES project is a consortium whose primary goal is to offer solutions to deal with distributed, volatile, and heterogeneous computing resources. The targeted platforms are virtually infinitely large, gathering an unbounded number of resources and services. Hence, SPADES' proposal includes i) an abstraction of heterogeneity through overlay computing systems; ii) a fully decentralized (peer-to-peer) distributed resource/service discovery system; iii) workflow scheduling techniques; iv) self-stabilizing mechanisms to tackle the unreliable nature of the platform; and finally v) optimizations regarding emerging Petascale architectures. To orchestrate such a platform, we have designed and implemented a novel middleware, called SBAM.

DLPT

The critical Service Discovery feature in SBAM is based on the Distributed Lexicographic Placement Table (DLPT). Relying on a prefix-tree distributed structure, DLPT offers required properties for the targeted environment such as dynamic load balancing or self-stabilization.

SBAM's DLPT is also available as a standalone component: ”Off The Shell DLPT”.

Download

Last Binary

From Sources

localhost:~ user$ svn checkout svn+ssh://graal.ens-lyon.fr/home/SVN/SPADES/trunk/SPADES/devel/sbam

Publications

Please use this reference to cite SBAM :

@article{FGCS2012,
  author    = {Eddy Caron and
               Florent Chuffart and 
               Cedric Tedeschi},
  title     = {{When Self-Stabilization Meets Real Platforms: an Experimental Study of a Peer-to-Peer Service Discovery System}},
  journal   = {Future Generation Comp. Syst.},
  url       = {http://dx.doi.org/10.1016/j.future.2012.10.003}
  volume    = {},
  number    = {},
  year      = {2012},
  pages     = {a paraitre},
}

@inproceedings{DLPT-SSS12,
	Title = {{O}ptimization in a {S}elf-{S}tabilizing {S}ervice {D}iscovery {F}ramework for {L}arge {S}cale {S}ystems},
	Author = {Caron, Eddy and Chuffart, Florent and Lamani, Anissa and Petit, Franck},
	Booktitle = {14th International Symposium on Stabilization, Safety, and Security of Distributed Systems (SSS 2012)},
	Editor = {W. Richa, Andr{\'e}a and Scheideler, Christian},
	Month = {October},
	Number = {7596},
	Pages = {239-252},
	Publisher = {Springer},
	Series = {Lecture Notes in Computer Science},
	Address = {Toronto, Canada},
	Year = {2012}
}

Demos

SBAM have been presented at SC10 and P2P2011.

http://graal.ens-lyon.fr/~fchuffart/files/sbam_demo_sc10.mov

Replay on your self CoPIF results (SSS 2012 à Toronto). Refer to the README file in the following archive

http://graal.ens-lyon.fr/~fchuffart/files/pifexp_realis.tgz

Other Involved Development Projects

 
sbam.txt · Last modified: 2012/10/12 09:41 by fchuffar · [Old revisions]
Recent changes RSS feed Powered by PHP Valid XHTML 1.0 Valid CSS Driven by DokuWiki