The Case for Interaction Paradigm Interoperability Georgios - PowerPoint PPT Presentation

Service-oriented Distributed Applications in the Future Internet: The Case for Interaction Paradigm Interoperability Georgios Bouloukakis 1 Joint work with Nikolaos Georgantas 1 , Sandrine Beauche 2 , Valérie Issarny 1 1 Inria Paris-Rocquencourt, France 2 Ambientic, France IFIP WG on Services-oriented Systems, Amsterdam, 11/10/2013

Service-oriented applications in the FI: Search & Rescue (S&R) Middleware life sign interoperability sensors required JavaSpaces data planning & JMEDS space deployment Web service S&R service S&R choreography personnel S&R coordinator Info diffusion & JMS coordination event-based system Heterogeneous Web Services middleware monitoring & JavaSpaces interaction control centers protocols JMS - 2

Interaction paradigms in middleware • Look for comprehensive systematic solution to middleware interoperability • Classify middleware implementations into families • Families follow well-known interaction paradigms Publish- Client-Server Interaction Tuple Space Subscribe paradigms (CS) (TS) (PS) Interaction Web Services JavaSpaces JMS protocols - 3

Some informal semantics of interaction paradigms Space coupling message(destination) 1.request-msg(destination) CS client server 2.reply-msg(source) Time coupling 1.subscription(topic) 2.event(topic) publishers subscribers publishers subscribers PS subscribers publishers broker 3.event(topic) Concurrency 1.tupl 2.tuple(template) publishers subscribers readers/take publishers subscribers e TS writers tspace rs - 4

State-of-the-art in interaction paradigm interoperability Theoretical approaches for individual paradigms • Rely on concurrency theory, process algebras, architectural connectors • No study of semantics across interconnected paradigms Enterprise Service Bus (ESB) Practical cross-paradigm approaches service A • Typically apply to specific middleware application implementations service B middleware Binding common bus Component protocol (BC) ESB - 5

Solution to middleware interoperability Generic Application (GA) Higher-level connector connector Interaction paradigm CS PS TS connectors Middleware platforms JavaSpaces Web Services Info diffusion & coordination PS connector Planning & connector CS deployment converter connector TS Sensing connector GA connector - 6

Mapping of space coupling – IDLs CS PS TS GA element message event tuple data main scope system ID system ID system ID system ID sub-scope operation topic template data qualifier interaction {one-way, {publish, subscribe} {write, take, read} {post, get, semantics notification, post-get, request-response, get-post} solicit-response} S&R element in: out: out: in/out data {sensorLocation, {personnelId, {sensorLocation, lifeSign} personnelLocation} lifeSign} out: {personnelLocation} - 7

Time coupling and concurrency of CS, PS, TS connectors • Specify connector protocols with Labeled Transition Systems PS connector • Express and verify semantics in LTL temporal logic fluent SUBSCRIBED = < {subscribe}, {unsubscribe} > assert EVENT_RECEIVED_IF_SUBSCRIBED_BEFORE_PUBLISH = forall [e:EVENT] [] ((SUBSCRIBED && publish[e]) -> (!event_lost[e] U get_next_return[e])) - 8 - 8

Mapping of time coupling and concurrency Info diffusion & coordination PS connector Planning & connector CS deployment converter connector TS Sensing connector protocol projection GA connector • Rely on the method of protocol conversion via projections 1 • Common semantics of CS, PS, TS apply to GA end-to-end - 9 1 S. S. Lam. Protocol Conversion. IEEE Trans. Softw. Eng. 14, 3. 1988.

Implementation of the GA connector: an eXtensible Service Bus (XSB) Planning & Info diffusion & Sensing application deployment coordination Web JavaSpaces middleware JMS Services • XSB is a generic bus BC BC BC BC • Build on top of a substrate bus Web middleware JavaSpaces JMS Services • Cross-paradigm integration in the S&R application CS/PS/TS TS CS PS GA GA GA GA common bus GA semantics protocol XSB on top of EasyESB 1 XSB - 10 1 https://research.linagora.com/display/easyesb

XSB architectural framework BC extensible architecture Refine to introduce a generic level new interaction From/to middleware paradigm Core Engine GA-IDL Processor Refine to introduce a interaction new middleware paradigm level Envelope for Substrate Bus middleware Refine to introduce a platform level new application service From/to substrate bus Refine to introduce a new substrate bus - 11

Evaluation – Effort for S&R application design Development effort for composing the three available application services into the S&R application CS-, PS-, TS-IDL Generated GA-IDL Data mapping descriptions descriptions directives (XML lines) (XML lines) (XML lines) All 3 application 407 249 226 services - 12

Evaluation – Extensibility of the XSB framework Development effort for the JMS binding component Lines of code PS-, GA-IDL schemas Configuration files (XML lines) (XML lines) Total effort 18021 2836 248 Developer’s effort 1162 191 12 Developer’s effort 6% 6% 4% ratio - 13

Evaluation - Performance • Latency overhead introduced by the bus for all interconnection combinations • Comparison between XSB and substrate bus Interconnection Latency overhead (ms) Comparison – CS-CS via EasyESB 258 CS-CS via XSB 261,5 +1,4% CS-PS via XSB 283 +9,7% CS-TS via XSB 276 +7,0% PS-TS via XSB 298 +15,5% - 14

Conclusion and future perspective • Service-oriented applications in the Future Internet • Require cross-middleware interoperability • Tackle this challenge via • Abstractions and mappings for interaction paradigms • Assessment of end-to-end semantics • Implementation into an extensible service bus  Next step • Extend with support for continuous interactions – data streaming protocols - 15

Thank you Further information: XSB : xsb.inria.fr CHOReOS : www.choreos.eu Inria ARLES : www.rocq.inria.fr/arles - 16