1. Dept. of Computer Science & Engineering, CUHK Fault Tolerance and Performance Analysis in Wireless CORBA Chen Xinyu 2002-12-09 Supervisor: Markers: Prof. Michael R. Lyu Prof. Jerome Yen Prof. John C.S. Lui

2. Outline u Motivation u Wireless CORBA u Fault Tolerant Wireless CORBA u Performance and Availability Analysis u Conclusions and Future Work

3. Motivation u Mobile Computing u Permanent failures Physical damage u Transient failures Mobile host Wireless link Environmental conditions u Fault Tolerant CORBA Entity replication

4. Visited Domain Home Domain Terminal Domain Wireless CORBA Architecture Access Bridge Static Host Terminal Bridge GIOP Tunnel ab 1 ab 2 mh 1 GTP Messages

5. Visited Domain ab 1 ab 2 Wireless CORBA Architecture Access Bridge Static Host Home Domain Home Location Agent Terminal Domain Terminal Bridge GIOP Tunnel mh 1 Terminal Domain Terminal Bridge GIOP Tunnel GIOP Tunnel mh 1 Terminal Domain Terminal Bridge GIOP Tunnel mh 1 Terminal Domain Terminal Bridge Access Bridge

6. Outline u Motivation u Wireless CORBA u Fault Tolerant Wireless CORBA u Performance and Availability Analysis u Conclusions and Future Work

7. Basic Concepts u Checkpoint the saved program’s states during failure-free execution u Repair brings the failed device back to normal operation u Rollback reloads the program’s states saved at the most recent checkpoint u Recovery the reprocessing of the program, starting from the most recent checkpoint, applying the logged messages and until the point just before the failure

8. Device, Wireless & Mobile Issues u Device Issues Slow processor Small memory Small disk space Low power supply Physical damage Applying mobile host as stable storage  a large number of system messages or a large size of information carried in one message   Checkpoints and Logs collection u Wireless Issues High bit error rate Little bandwidth Long transfer delay u Mobile Issue Handoff  Applying Access Bridge as stable storage  Uncoordinated checkpointing Pessimistic message logging

9. Fault Tolerance Architecture Client Object Terminal Bridge Recovery Mechanism ORB Platform Mobile Host Recovery Mechanism Logging Mechanism Platform Access Bridge Mobile Side Fixed Side Mobile Support Station ORB Recovery Mechanism Logging Mechanism ORB Platform Static Server GIOP Tunnel Multicast Messages Object Replica

10. Mobile Host Handoff Access Bridge 1 Access Bridge 2 Access Bridge 3 Home Location Agent Handoff Location Update

11. Home Location Agent Mobile Host Handoff Access Bridge 1 Access Bridge 2 Access Bridge 3 Handoff Location Update

12. Home Location Agent Mobile Host Crash Access Bridge 1 Access Bridge 2 Access Bridge 3

13. Home Location Agent Mobile Host Recovery Access Bridge 1 Access Bridge 2 Access Bridge 3 Collect last checkpoint and succeeded message logs Sorted by Ack. SN Reconnect Messages Replay

14. Outline u Motivation u Wireless CORBA u Fault Tolerant Wireless CORBA u Performance and Availability Analysis u Conclusions and Future Work

15. Assumptions u Failure occurrence, message arrival and handoff event  homogeneous Poisson process with parameter,  and  respectively u Failures do not occur when the program is in the repair or rollback process u A failure is detected as soon as it occurs

16. Execution without Checkpointing RY0Y0 X0X0 R F1F1 H1H1 Z0Z0 0 t FjFj HkHk m j (1)m j (N)m 1 (n 1 )m 0 (N) X(N) RepairHandoff HH

17. Conditional Execution Time & LST

18. LST and Expectation of Program Execution Time u u

19. CiCi Execution with Equi-number Checkpointing R+C Y i (0) X i (0) R+C F i (1) H i (1) Z i (0) 0 t F i (j) H i (k) m ij (1)m ij (a)m i1 (n i1 )m i0 (a) X i (N,a) Repair + RollbackHandoff C i-1 Checkpointing HHCC

20. Conditional Execution Time & LST

21. LST and Expectation of Program Execution Time u u

22. Average Availability u uptime interval: a program produces useful work towards its completion u downtime interval: Repair and rollback Handoff Checkpoint creation Wasted Computation u average availability: how much of the time an MH is in uptime interval during an execution

23. Optimal Checkpointing Interval u 

24. Beneficial Condition u 

25. Equi-number Checkpointing u Equi-number checkpointing with respect to message number Message number in each checkpointing interval is not changed u Equi-number checkpointing with respect to checkpoint number Checkpoint number is not changed

26. Equi-number Checkpointing with respect to Checkpoint Number

27. Equi-number Checkpointing with respect to Message Number

28. Comparison Between Checkpointing and Without Checkpointing

29. Average Availability vs. Message Arrival Rate and Handoff Rate

30. Conclusions u Fault tolerant wireless CORBA u Equi-number checkpoiting strategy u LST and expectation of program execution time u Average availability u Optimal checkpointing interval u Beneficial condition

31. Future Work u Analysis model The message queuing effect during repair and recovery u Failure detector Distributed consensus with link failures, process failures, and mobile disconnections Leads to a faster solution Reduces communication costs u Fault tolerance in Ad Hoc network Without infrastructure support Self-organizing and adaptive

33. Thank You