1 / 26

AMPol-Q : A daptive M iddleware Pol icy to support Q oS

AMPol-Q : A daptive M iddleware Pol icy to support Q oS. Raja Afandi , Jianqing Zhang, Carl A. Gunter Computer Science Department, University of Illinois Urbana-Champaign. QoS Based Discovery. Services are discovered and selected on the basis of functional properties (IOPE)

isolde
Download Presentation

AMPol-Q : A daptive M iddleware Pol icy to support Q oS

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. AMPol-Q: Adaptive Middleware Policy to support QoS Raja Afandi, Jianqing Zhang, Carl A. Gunter Computer Science Department, University of Illinois Urbana-Champaign

  2. QoS Based Discovery • Services are discovered and selected on the basis of functional properties (IOPE) • Difficult to discover and compose services on the basis of non-functional features (QoS) • Complex to model QoS features • Global nature of QoS • QoS composition requires complex calculations of aggregate and global QoS values

  3. Global QoS • Client policies constraining Service A & B • Message Sender requires the message to be confidential • Business client needs its loan request to be processed with in a week • End service policies constraining Client • Bank requires applicant credit score to be fair • RMTA requires that a message from a sender must not be greater than 10KB • Global QoS • End-to-End QoS • Two-way QoS constraints Service A Service B Client • Client & Service A • Sender MUA and SMTA • Business Client and Loan Agency • Service A & Service B • SMTA and RMTA • Loan Agency & Bank

  4. Global QoS: Discovery • Discovery approaches: • Limited to discovering first-level immediate services • Each individual service is responsible for discovering other services independently • Global QoS evaluation requires global information of all the entities (Global View) Service B3 ? Service A Service B2 Client Service B1

  5. QoS Policies • Supporting QoS also requires monitoring and enforcing QoS behavior and policies • Support for non-functional features such as security and reliability breaks the interoperability of the system • Diverse Policies • Constraints may change

  6. Main Objectives… • Global QoS and dynamic discovery • QoS Policies and Interoperability

  7. Strategy • End-to-End (E2E) integrated approach • Description: Semantic modeling of QoS capabilities and constraints (policy rules) • Global Discovery: Global service discovery and QoS analysis • Distribute Monitoring: Semantic model and distributed enforcement • Two-Way-Specialization: entities can dynamically adapt to the advertised requirements of each other • Basic architecture based on three frameworks • Description framework defines QoS model and describes declarative domain-specific policy rules • Discovery framework governs how to publish, find, and select services on the basis of QoS • Monitoring framework provides means to adhere to and enforce published QoS requirements and constraints

  8. Case Study: Messaging System • WSEmail • Internet messaging based on web services SS RS SS: Sender Mail Server RS: Recipient Mail Server SC RC

  9. Description Framework • Includes semantic models for describing QoS capabilities, constraints and requirements of different entities in the system: • QoS Model • Policy Model • Entity Profile Model • Messaging Domain Model

  10. Description Framework: QoS Model APES QoS Domain Ontology QoS Base Ontology QoS Monitoring Ontology

  11. Description Framework : Policy Model • Semantic language of constraints and requirements • Rules are defined using QoS ontology and SWRL rule format: ampol:Encryption(?a) ∧ ampol:algoType(?a, ?b) ∧ ampol:stringValue(?b, "IBE2.3") → ampol:verified(RC_Enc_Rule1, true) ampol:Attachment(?a) ∧ ampol:attachmentSize(?a, ?b) ∧ ampol:intValue(?b, ?c) ∧ swrlb:lessThanOrEqual(?c, 1000) ∧ ampol:unit(?b, "KB") → ampol:verified(RS_Att_Rule1, true) • Supports • Delegation of policy enforcement • Rule prioritization to resolve conflicts • Public vs. private rules

  12. Description Framework : Policy Model

  13. Description Framework : Entity Profile Model • Type of service description profile • Associated with a system entity and can be advertised with OWL-S service profile • Profiles represent entity : • Domain vocabulary • Capabilities • Constraints • Extension constraints • Service dependencies • OWL-S request templates • Clients uses it to discover desired services • Supports global discovery better than current approaches (based on pure OWL-S, UDDI and IOPE) Sample Entity Profile Sample Service Request Sample Service Profile

  14. Discovery Framework • Discover global QoS information about all the candidate services • Select best matched services • Binding selected parties in a QoS contract • Consist of three components: • Service Discovery and Chaining • Global QoS Analysis • Policy Agreement and Contract Negotiation

  15. Discovery: Service Chain Graph Request Template Request Template B CR Dependency List Dependency List CR Request Template Dependency List Dependency List

  16. Global QoS Analysis CB1 CB1 FL1 FB1 FCR1 FC1 C1 L1 B1 CR1 RC1 RC1 RC1 RC1 CB1  FC1 R = Requirements C = Constraints F = Capabilities E = Extension Capabilities EP = Extension Policy CB1  E EPC1 Agreement  Agreement Value No Agreement  Penalty

  17. Monitoring Framework • Monitoring involves measuring delivered QoS, verifying QoS constraints and taking enforcement actions • QoS feature is associated with Monitoring Processes: • Measurement • Enforcement • Adherence • Policy Framework control the system QoS policies • Adapts by adding or replacing pluggable components • Policy conformance and enforcement logic as extensions • Generic enough to process any complex constraints • Monitoring processes are implemented as pluggable extensions • Extension Manager • Policies control the download and execution of extensions • Download plug-ins from secure third-party plug-in server

  18. AMPol-Q Middleware

  19. AMPol-Q Middleware Implementation

  20. Case Study: Policy Based WSEmail • WSEmail: Web services based Internet messaging • Integrated AMPol-Q middleware with WSEmail • Transparent Integration • Implemented APES features: • Payment as Puzzle • Hashcash • RTT • Encryption using Identity-Based Encryption (IBE)

  21. Case Study: Overview Server Ingress Policies • HashCash Puzzle RMTA (RS) Sandy SMTA (SS) Gary Server Egress Policies • Attachment Size Entity Profiles Registry Server Client Ingress Policies • IBE Encryption Recipient MUA (RC) afandisandy@sandy Sender MUA (SC) afandigary@gary Plugin Server

  22. AMPol-Q Contributions • End-to-end solution for supporting non-functional constraints, capabilities & requirements • Comprehensive semantic QoS Model • Global service discovery and global QoS analysis • Adaptive policy framework for QoS monitoring • Reference architecture for adaptive middleware for messaging systems • Validation of proposed approach through a case study on WSEmail • One of the most complete studies to date of a proof-of-concept QoS-aware policy system based on Web services

  23. Current and Future Work • Formal security analysis and Improved security measures • Policy conflict resolution and models for negotiation • Performance testing • Inter-Ontology relationship for heterogeneous ontologies

  24. Questions… • Thanks…

  25. Discovery: Sample Merged Policy SC adherence policy SC Enforcement policy SS Enforcement policy RS Enforcement policy Delegated Rules RC Enforcement policy

  26. Query Matching • Service query matching constraints

More Related