1. 1 Building Reliable Web Services: Methodology, Composition, Modeling and Experiment Pat. P. W. Chan Supervised by Michael R. Lyu Department of Computer Science and Engineering The Chinese University of Hong Kong 29 November 2007

2. 2 Outline  Introduction Contribution Related Work Web Services  Problem Statement  Methodologies for Web Service Reliability  New Reliable Web Service Paradigm Optimal Parameters Experimental Results  Web Service Composition Algorithm Experimental Results Discussion  Modeling of the Paradigm  Conclusion and Future Work

3. 3 Introduction  Service-oriented computing is becoming a reality.  Web Service is a promoting technique in the internet.  The benefit of interoperability, reusability, and adaptability.  The problems of service dependability, security and timeliness are becoming critical.  Reliability is an important issue.  Existing web service model needs to be extended to assure reliability.  We propose experimental settings and offer a roadmap to dependable Web services. Introduction

4. 4 Contribution  Surveyed on reliability methodologies  Surveyed on Web services reliability and Web service composition techniques  Proposed an architecture for dependable Web services  Proposed an algorithm for Web services composition  Developed reliability models for the proposed scheme  Performed experiments for evaluating the reliability of the system and the correctness of the algorithm Introduction

5. 5 Reliability  "a measure of the success with which the system conforms to some authoritative specification" Guaranteed delivery Duplicate elimination Ordering Crash tolerance State synchronization Introduction

6. 6 What are Web Services ?  Self-contained, modular applications built on deployed network infrastructure including XML and HTTP  Use open standards for description (WSDL), discovery (UDDI) and invocation (SOAP) Introduction

7. 7 Web Services Internet UDDI WSDL HTTP/SOAP WSDL Introduction

8. 8 Web Services Architecture SOAP HTTP/SMTPXMLTCP/IP Directory Inspection Building Block Modules Building Block Modules Inter Application Protocols Referral Routing Security License EventingTransactions Reliable Messaging The Internet Description … … Introduction

9. 9 Web Services  Benefits of WS Service-oriented Highly accessible Open specification Easy integration  Number of system using Web service including: shopping, e-banking … Simplicity DynamicStandard Web Services Build common infrastructure reducing the barriers of business integration with lower costs and faster speed. Introduction

10. 10 Problems of Web Services  Transaction Atomicity is not provided  Security Insecure Internet transportation  Reliability The internet is inherently unreliable No single underlying “ transport protocols ” address all the reliability issues. Introduction

11. 11 Problem Statement  Fault-tolerant techniques Replication Diversity  Replication is one of the efficient ways for providing reliable systems by time or space redundancy. Increasing the availability of distributed systems Key components are re-executed or replicated Protect against hardware malfunctions or transient system faults  Another efficient technique is design diversity Employ independently designed software systems or services with different programming teams, Defend against permanent software design faults.  We focus on the analysis of the replication techniques when applied to Web services.  A generic Web service system with spatial as well as temporal replication is proposed and investigated. Introduction

12. 12 Road Map for Research  Redundancy in time Retry Reboot  Redundancy in space Sequentially Parallel Majority voting using N modular redundancy Diversified version of different services Introduction

13. 13 Proposed Paradigm Replication Manager RR Algorithm / Voting WatchDog UDDI Registry WSDL Web Service IIS Application Database Web Service IIS Application Database Web Service IIS Application Database Client Port Application Database Register Look up Get WSDL Invoke web service Keep check the availability of all the web. If Web service failed, update the list of availability of Web services Update the WSDL Reliable Web Service Paradigm

14. 14 Different Approaches  Replication Round-robin scheduling algorithm  Design Diversity N-version programming Recovery block Reliable Web Service Paradigm

15. 15 Replication: Round-robin Reliable Web Service Paradigm

16. 16 RM sends message to the Web Service Update the Web service availability list Do not get reply Map the new address to the WSDL after RR System Fail Get reply All Service failed Work Flow of the Replication Manager Reliable Web Service Paradigm

17. 17 Design Diversity: Parallel N-Version Programming Reliable Web Service Paradigm

18. 18 Design Diversity: Recovery Block Reliable Web Service Paradigm

19. 19 Experiments Variations  A series of experiments are designed and performed for evaluating the reliability of the Web service. Spatial replication00001111 Reboot00110011 Retry01010101 Reliable Web Service Paradigm 12345678

20. 20 Varying the parameters  Number of tries  Timeout period for retry in single server  Timeout period for retry in our paradigm  Polling frequency  Number of replicas  Load of server Reliable Web Service Paradigm

21. 21 Number of tries Number of failures in Temp Number of failures in Perm 09576 122 200 300 400 500 Reliable Web Service Paradigm

22. 22 Timeout period for retry in single server Timeout period for retry (s) Number of failures in Temp Number of failures in Perm 0957265 227156 507314 606890 70189 8082 9011 1002 1200 1400 1600 1800 Reliable Web Service Paradigm

23. 23 Timeout period for retry in single server # of failure Timeout period Reliable Web Service Paradigm

24. 24 Timeout period for retry in our paradigm Timeout period for retry (s) Number of failures in Temp Number of failures in Perm 0281 202 500 1000 2000 Reliable Web Service Paradigm

25. 25 Polling frequency (number of requests per min) Number of failures in Temp Number of failures in Perm 007124 10811 2030 5012 1001 15213254 2011241023 Reliable Web Service Paradigm

26. 26 Polling frequency # of failure Polling frequency Reliable Web Service Paradigm

27. 27 Number of Replicas Number of replicasNumber of failures in TempNumber of failures in Perm No replica918152 22356 300 400 Reliable Web Service Paradigm

28. 28 Load of Web Server Load of the web server (%) Number of failures in Temp Number of failures in Perm 7000 7500 8023 851014 90512528 9532143125 9887928845 9989978994 Reliable Web Service Paradigm

29. 29 Summary of Parameters  Number of tries = 2  Timeout period for retry in single server = 10s  Timeout period for retry in our paradigm = 5s  Polling frequency = 10 request per min  Number of replicas = 3  Load of server < 75% Reliable Web Service Paradigm

30. 30 Testing system  Best Route Finding.  Provide traveling suggestions for users.  Starting point and destination.  The system needs to provide the best route and the price for the users. Reliable Web Service Paradigm

31. 31 System Architecture Reliable Web Service Paradigm

32. 32 Experimental Setup  Examine the computation to communication ratio  Examine the request frequency to limit the load of the server to 75%  Fix the following parameters Computation to communication ratio (e.g 10:1) Request frequency Reliable Web Service Paradigm

33. 33 Experimental Setup Communication time: Computation time 143:14 (10:1) Request frequency1 request per min Load78.5% Timeout period of retry1 min Timeout for Web service in RM1s (web service specific) Polling frequency10 requests per min Number of replicas5 Max number of retries5 Round-robin rate1 s Reliable Web Service Paradigm

34. 34 Experiment Parameters  Fault mode Temporary (fault probability: 0.01) Permanent (fault probability: 0.001)  Experiment time 5 days (7200 requests)  Measure: Number of failures Average response time (ms)  Failure definition: 5 retries are allowed. If there is still no correct result from the Web service after 5 retries, it is considered as a failure. Reliable Web Service Paradigm

35. 35 Experimental Result with Round-robin (failures / response time in ms) ExperimentsSingle serverSingle server with retry Single server with reboot (continues no response for 3 requests) Single server with retry and reboot Spatial Replication RR Hybrid approach RR+Retry Hybrid approach RR+ Reboot All round approach RR spatial + Retry (5 times) + Reboots Normal case0 / 1830 / 1930 / 1900 / 1870 / 1880 / 1950 / 1930 / 190 Temp705 / 1900 / 223723 / 2310 / 238711 / 1870 / 233726 / 1880 / 231 Perm6144 / --6337 / --1064 / --5 / 25785637 / --5532 / --152 / 1870 / 191 Reliable Web Service Paradigm 12345678

36. 36 Experimental Result with N-Version (failures / response time in ms) ExperimentsSingle serverSingle server with retry Single server with reboot (continues no response for 3 requests) Single server with retry and reboot Spatial Replication Voting Hybrid approach Voting+ Retry Hybrid approach Voting + Reboot All round approach Voting spatial + Retry (5 times) + reboots Normal case0 / 1830 / 1930 / 1900 / 1870 / 1890 / 1900 / 188 Temp705 / 1900 / 223723 / 2310 / 2380 / 190 0 / 1890 / 187 Perm6144 / --6337 / --1064 / --5 / 25783125 / 1913418 / 19240 / 1890 / 188 Reliable Web Service Paradigm 12345678

37. 37 Experimental Result with Recovery Block (failures / response time in ms) ExperimentsSingle serverSingle server with retry Single server with reboot (continues no response for 3 requests) Single server with retry and reboot Spatial Replication Voting Hybrid approach Voting+ Retry Hybrid approach rollback + Reboot All round approach rollback spatial + Retry (5 times) + reboots Normal case0 / 1830 / 1930 / 1900 / 1870 / 1910 / 1890 / 1930 / 188 Temp705 / 1900 / 223723 / 2310 / 2380 / 2050 / 2030 / 2040 / 201 Perm6144 / --6337 / --1064 / --5 / 25783478 / 2153245 / 208201 / 2110 / 201 Reliable Web Service Paradigm 12345678

38. 38 Summary of the proposed paradigm  Temporal replication improves the reliability.  Spatial replication further improves the reliability of Web services.  N-version programming approach is the most reliable and efficient. Reliable Web Service Paradigm

39. 39 Web Service Composition Algorithm  N-version programming Reliable Efficient  Composition WSDL – Web Services Description Language WSCI – Web Services Choreography Interface  Verification BPEL – Business Process Execution Language Petri-Net Web Service Composition

40. 40 … … WSDL Web Service Composition

41. 41 <action name="ReceiveStartpointDest “ role="tns:busAgent “ operation="tns:BRF/shortestpath"> <action name="Receivecheckpoint “ role=" tns:busAgent “ operation="tns:BRF/addCheckpoint"> … WSCI Web Service Composition

42. 42 Web Service Composition

43. 43 1.Output 2.Operation in WSDL 3.Find the output information in CP1 (Web service component) 4.If Input of the operation == required input CP1 5.Else search in the WSCI of CP1 to find action == operation 6.Get the pervious action involved 7.Search in WSDL to find operation == action 8.If Input of the operation == required input CP4 Else, till the root of WSCI Web service composition CP7 CP10 Web Service Composition

44. 44 Web Service Composition Search Agent Bus AgentTrain Agent Bus: KMB Train: MTR Starting P1 Starting P2 Web Service Composed Tree

45. 45 Web Service Composition

46. 46 Petri-Net – Basic Activities Web Service Composition

47. 47 Petri-Net – Structure Activities Web Service Composition

48. 48 Composed Petri-Net Web Service Composition

49. 49 Web Service Composition

50. 50 Web Service Composition

51. 51 Summary of the Web Service Composition Algorithm  The composition algorithm is proposed with the use of WSDL and WSCL  The BPEL of the composed Web services are generated  Petri-Net is employed to avoid deadlock  Acceptance tests are set for checking the correctness  Experiments are performed Efficient Accurate Deadlock-free Web Service Composition

52. 52 Experimental Setup  Same as the previous setting  Employ the composed Web services (BRF)  Fault Injection Temporary Permanent Byzantine failure Network failure Web Service Composition

53. 53 Experimental Result (1) Web Service Composition

54. 54 Experimental Result (2) Web Service Composition

55. 55 Experimental Result (3) Web Service Composition

56. 56 Experimental Result (4) Web Service Composition

57. 57 Modeling  Modeling can check the reliability, correctness, deadlock-free and performance of the system  We employed Petri-Net Markov chain model Web Service Composition

58. 58 Petri-Net (Four identical replicas) Modeling

59. 59 Petri-Net (N-version Web service with voting) Modeling

60. 60 Petri-Net (Recovery Block) Modeling

61. 61 (a) (b) P1P1 λ1λ1 μ1c2μ1c2 S-j P2P2 μ2c2μ2c2 λ2λ2 S-j-1 S S-n F λNλN μ*c 2 (1-c 1 )μ* λ*λ* S-1S-2 λ*λ* μ*c2μ*c2 λ*λ* (1-c 1 )μ* F (1-c 1 )μ 1 (1-c 1 )μ 2 (1-c 2 )μ 1 (1-c 2 )μ 2 Reliability Model Modeling

62. 62 Reliability Model IDDescriptionValue λnλn Network failure rate0.02 λ*Web service failure rate0.025 λ1λ1 Resource problem rate0.142 λ2λ2 Entry point failure rate0.150 μ*Web service repair rate0.286 μ1μ1 Resource problem repair rate0.979 μ2μ2 Entry point failure repair rate0.979 C1C1 Probability that the RM response on time0.9 C2C2 Probability that the server reboot successfully0.9 Modeling

63. 63 Outcome (SHARPE) Modeling

64. 64 Conclusion  Surveyed replication and design diversity techniques for reliable services and the state-of-the-art Web service composition algorithm.  Proposed a hybrid approach to improving the reliability of Web services.  Optimal parameters are obtained.  Proposed a Web service composition algorithm and verified by Petri-Net.  Carried out a series of experiments to evaluate the availability and reliability of the proposed Web service system.  Employ Petri-Net and Markov chain to model the system to analysis the reliability and performance. Conclusion and Future Work

65. 65 Future Work  Improve the current fault-tolerant techniques Current approach can deal with hardware and software failures. How about software fault detectors?  N-version programming Different providers provide different solutions. There is a problem in failover or switch between the Web Services.  Application Different requirements Realize in the Internet. Conclusion and Future Work

66. 66 Q&A