B. Yao, K.-F. Ssu, and W. K. Fuchs. Message logging in mobile computing. In 29th IEEE Int. Symp. on Fault-Tolerant Computing (FTCS-29), pages 294--301. IEEE CS Press, 1999.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....the system failures; and after the rollback, the process can immediately resume its computation without waiting for any coordination message from another process. For the mobile computing environment, some asynchronous recovery schemes based on the pessimistic message logging have been proposed in [20, 21, 22, 32]. However, the pessimistic message logging may be the less plausible choice considering the relatively high logging overhead for the frequent access to the stable storage. One alternative to reduce the storage access overhead is to use the causal logging scheme [3, 4, 12] However, in this ....

B. Yao, K. Ssu, and W.K. Fuchs, "Message Logging in Mobile Computing," Proc. 29th Symp. on Fault Tolerant Computing Systems, 1999, pp. 294--301. 35

....failure. Also, after the rollback, the process can immediately resume its computation without waiting for any coordination message from another process. For the mobile environment, some asynchronous recovery schemes based on the pessimistic message logging (denoted by PML) have been proposed in [15, 16, 17, 25]. However, the pessimistic logging may be the less plausible choice considering the frequent access to the stable storage. To avoid the frequent stable storage access, a causal logging scheme (denoted by CML) 3, 8] can be used. However, in this scheme, a large size of dependency information ....

B. Yao, K. Ssu, and W.K. Fuchs, "Message Logging in Mobile Computing," Proc. 29th Symp. on Fault Tolerant Computing Systems, 1999, pp. 294--301. 34

.... overhead and a little inference with MHs has been proposed in [17] Also, in spite of the heavy stable storage access overhead, pessimistic logging is considered as a good design choice for the mobile environment, because of its simple implementation and the capability of asynchronous recovery [19, 28]. In the mobile computing environment, the MHs are frequently disconnected from the network with out a failure and the exchange of a large number of coordination messages is considered too costly. Hence, recovery coordination among the MHs is not plausible and asynchronous recovery must be ....

....and logs no longer required for any recovery must be discarded for the reuse of the space. Such garbage collection may have to put extra communication overhead if the logs are dispersed over a large number of cells. Some works related to the distributed storage management have been proposed in [19, 28]. For the fast recovery, it is desirable for checkpoints and message logs to be near the MH on recovery, and hence, checkpoints and logs of a MH in [19] keep moving as the MH performs the handoff between two cells. As a result, instant recovery can be possible, however, the failure free ....

[Article contains additional citation context not shown here]

B. Yao, K. Ssu and W.K. Fuchs. Message Logging in Mobile Computing. In Proc. of the 29th Syrup. on Fault Tolerant Computing Systems, pp. 294-301, Jun. 1999. 21

.... include [6] 7] applications) 8] 9] workflows) For mobile systems specifically, work includes [10] 11] transactional support) 12] 13] 14] database consistency) 15] 16] concurrency control and correctness) The scenarios in this paper follow [3] mobile recovery) and [17] (logging messages) There is little work on formalizing recovery [18] 19] and it does not address introducing abstraction. In this paper, we discussed and specified mobile system recovery, showing how operations by a mobile host are recoverable under different strategies. The variants depend ....

B. Yao, K.-F. Ssu, and W.K. Fuchs, "Message Logging in Mobile Computing," Proc. Symp. Fault-Tolerant Computing, pp. 294-301, 1999.

....in addition to the independent checkpointing [10] For the logging, a pessimistic, optimistic [11] or causal scheme [12] can be used. For the mobile systems, an optimistic logging scheme with the low cost has been proposed in [13] and some pessimistic logging schemes also have been considered in [14,15,16]. In most of the logging schemes, MSSs are used to manage checkpoints and message logs of a MH, due to the lack of the stable storage of the MH. Hence, the recovery information of a MH becomes dispersed over a number of MSSs, as the MH moves around the cells. When a failure occurs, the MH should ....

....recovery from a failure is very important, and hence, the efficient management of distributed recovery information becomes an important design issue for the checkpointing recovery. To cope with this problem, a MH in [15] carries checkpoints and message logs as it moves, and the scheme suggested in [16] utilizes the home of each MH as a centralized recovery information manager. The schemes in [14] use the movement based recovery information transfer. In this paper, an efficient distributed storage management scheme is proposed for the message logging in mobile environments. In the proposed ....

[Article contains additional citation context not shown here]

B. Yao, K. Ssu & W.K. Fuchs, Message logging in mobile computing, Proc. 29th Symp. on Fault Tolerant Computing Systems, Madison, Wisconsin, 1999, 294-301.

....extended schemes of coordinated check pointing, communication pattern based checkpointing and communication induced checkpointing have been proposed in [5] 1] and [2] respectively. Also, for the asynchronous recovery, optimistic and pessimistic logging schemes are considered in [3] and in [4, 6], respectively. To save checkpoints and message logs of a MH, the stable storage of mobile support stations(MSSs) are used due to the lack of the spaces in MHs. Hence, as the MH moves, the recovery information of the MH becomes dispersed over a number of MSSs, and in case of a failure, the MH ....

....of MSSs. Considering the cost to transfer the recovery information during the hand off or to collect it during the recovery, efficient management of distributed information becomes an important issue to design the fault tolerant mobile system. Some suggestions for this problem have been made in [4, 6]. For fast recovery, a MH in [4] carries the recovery informa tion as it moves, and in [6] the home of each MH is utilized as a centralized information storage. This paper presents distributed storage management schemes based on a region manager for efficient management of recovery information ....

[Article contains additional citation context not shown here]

B. Yao, K. Ssu and W.K. Fuchs. Message logging in mobile computing. In Proc. of the 29th Symp. on Fault Tolerant Computing Systems, pages 294301, 1999.

.... overhead and a little inference with MHs has been proposed in [17] Also, in spite of the heavy stable storage access overhead, pessimistic logging is considered as a good design choice for the mobile environment, because of its simple implementation and the capability of asynchronous recovery [18, 27]. In mobile computing environment, the MHs are frequently disconnected from the network without a failure and the exchange of a large number of coordination messages is considered too costly. Hence, recovery coordination among the MHs is not plausible and asynchronous recovery must be sought. ....

....and logs no longer required for any recovery must be discarded for the reuse of space. Such garbage collection may have to put extra communication overhead if the logs are dispersed over a large number of cells. Some works related to the distributed storage management have been proposed in [18, 27]. For fast recovery, it is desirable the checkpoints and message logs to be near the MH on recovery, and hence, checkpoints and logs of a MH in [18] keep moving as the MH performs the handoff between cells. As a result, instant recovery can be possible, however, the failure free communication ....

[Article contains additional citation context not shown here]

B. Yao, K. Ssu and W.K. Fuchs, "Message logging in mobile computing," In Proc. of the 29th Syrup. on Fault Tolerant Computing Systems, Jun. 1999. 21

....orphan processes in spite of base station failures and achieves run time performance similar to that of asynchronous logging. 1 Introduction Messages logging at base stations or WAP gateways to provide failure recovery for processes on mobile hosts has been studied in several research projects [1, 13, 7, 4, 3]. As messages are logged at base stations, mobile hosts are not required to have stable storage. If receiver based message logging is implemented, message logs can be sent to recovering processes by base stations during recovery, instead of by other mobile hosts. There is no need for a recovering ....

....base stations during recovery, instead of by other mobile hosts. There is no need for a recovering process to send messages to other mobile hosts and wait for replies. Using this approach, processes on mobile hosts can quickly and independently recover, regardless of failures of other mobile hosts [4]. This is in contrast to consistent checkpoint protocols where multiple processes on different mobile hosts have to roll back to recover from a single failure. Two alternative message logging approaches are typically possible: synchronous and asynchronous logging. In asynchronous logging, the ....

[Article contains additional citation context not shown here]

B. Yao, K. Ssu and W.K. Fuchs. Message Logging in Mobile Computing. In Proceedings of the Twenty-Ninth Annual International Symposium on Fault-Tolerant Computing, pages 294--301, June 1999.

....mechanisms for client state tracking and failure recovery. Experiment results on the latency, throughput and migration latency of the implementation are provided. 2 Related Work Using network layer agents or application layer proxies to log messages has been studied in previous research projects [17, 7, 14, 10, 4]. These protocols treat message logs and checkpoints as byte blocks and do not attempt to extract from them information that can aid client recovery. An optimization technique that provides logging bandwidth similar to optimistic logging and maintains the semantics of pessimistic logging has been ....

B. Yao, K. Ssu and W.K. Fuchs. Message Logging in Mobile Computing. In Proceedings of the Twenty-Ninth Annual International Symposium on Fault-Tolerant Computing, pages 294--301, June 1999.

....WavePOINT II to communicate with the PCs. The workstation was connected to a file server, a dual processor Sun Ultra Sparc 2 with 512MB RAM running Solaris 2.6, by a 155Mbps ATM network. An independent checkpointing protocol with receiverbased pessimistic logging was used for our experiments [4]. There were two processes simulating web browsing behavior using pre generated traces 1 on each mobile host. The processes sent requests and replies to each other. Each message was sent to the mobile support station through the wireless channel and then forwarded to the destination. The mobile ....

....and communication patterns. In contrast, the fixed checkpoint interval approach did not accurately bound the recovery time. With a five minute fixed checkpoint interval, the maximum recovery time ranged from 49 seconds to 119 seconds. 1 The details for generating the traces can be found in [4]. Table 2: Maximum Recovery Time Trace Fixed Interval Our Checkpoint Mechanism Program Recovery Time Recovery Time (sec) sec) T1 119.3 306.0 T2 73.3 303.7 T3 48.7 297.3 Table 3: Comparison of Execution Time Trace Normal Fixed Interval Our Mechanism Exec Exec No. of Exec No. of Time (s) ....

B. Yao, K. Ssu, and W. K. Fuchs, "Message Logging in Mobile Computing," Proceedings of IEEE Fault-Tolerant Computing Symposium, June 1999.

....battery power, and host mobility [2 4] It is not appropriate to directly apply many of the checkpointing and recovery protocols [5 11] designed for fixed network distributed systems to mobile environments. Several checkpointing protocols for wireless mobile environments have been proposed [12 18]. These protocols generally require the availability of extensive stable storage. Because storage on the mobile host is typically not considered stable, most of these protocols store checkpoints and message logs on local base stations. Stable storage on the base station is also used to keep ....

....and logs all messages received. Both protocols suggest saving checkpoints in the stable storage on the base stations instead of on the mobile hosts. Yao, et al. developed an approach to independent checkpointing with receiver based logging for fast recovery and efficient garbage collection [18]. Prakash and Singhal developed a non blocking coordinated checkpointing protocol that requires a minimum number of mobile hosts to participate in checkpointing [14] Cao and Singhal showed that 4 the protocol may result in inconsistent global states that cannot be used for recovery [16] The ....

B. Yao, K.F. Ssu, W.K. Fuchs, "Message logging in mobile computing", Proceedings of IEEE Fault-Tolerant Computing Symposium, 1999 June, pp 294 -- 301.

....battery power, and host mobility [2 4] It is not appropriate to directly apply many of the checkpointing and recovery protocols [5 11] designed for fixed network distributed systems to mobile environments. Several checkpointing protocols for wireless mobile environments have been proposed [12 18]. These protocols generally require the availability of extensive stable storage. Because storage on the mobile host is typically not considered stable, most of these protocols store checkpoints and message logs on local base stations. Stable storage on the base station is also used to keep ....

....and logs all messages received. Both protocols suggest saving checkpoints in the stable storage on the base stations instead of on the mobile hosts. Yao, et al. developed an approach to independent checkpointing with receiver based logging for fast recovery and efficient garbage collection [18]. Prakash and Singhal developed a non blocking coordinated checkpointing protocol that requires only a minimum number of mobile hosts to participate in checkpointing [14] Cao and Singhal showed 4 that the protocol may result in inconsistent global states that cannot be used for recovery [16] ....

B. Yao, K.F. Ssu, W.K. Fuchs, "Message logging in mobile computing", Proceedings of IEEE Fault-Tolerant Computing Symposium, 1999 June, pp 294 -- 301.

.... message logging applications can be much smaller than the fixed checkpoint interval, especially when the communication to computation ratio is high [16 18] Figure 1 shows experimental results for recovery times 1 on a distributed shared memory (DSM) system and a wireless local area network [16, 17]. With message logging, the application saves every incoming and or outgoing message in addition to checkpointing periodically [19 21] During recovery, the logged messages are immediately available and do not have to be reproduced. Programs will not be delayed by waiting for results from other ....

....server and a file server, respectively. All devices were connected by a 100Mbps Ethernet. TCP IP communication was used for both environments. 4. 2 Message logging protocol The adaptive checkpointing protocol with receiverbased pessimistic message logging was implemented for the experiments [17]. The reliable FIFO channel, the fail stop fault model, and piecewise determinism were assumed in the message logging protocol. In the wireless environment, two processes simulated web browsing behavior using five pre generated traces with varying communication rates on each mobile host. The ....

B. Yao, K. F. Ssu, and W. K. Fuchs, "Message Logging in Mobile Computing," Proceedings of IEEE Fault-Tolerant Computing Symposium, pp. 294--301, June 1999.

No context found.

B. Yao, K.-F. Ssu, and W. K. Fuchs. Message logging in mobile computing. In 29th IEEE Int. Symp. on Fault-Tolerant Computing (FTCS-29), pages 294--301. IEEE CS Press, 1999.

No context found.

Yao, B., Ssu, K. F., and Fuchs, W. K. Message logging in mobile computing. In Proceedings of the IEEE Fault-Tolerant Computing Symposium (June 1999), pp. 294-301.

No context found.

B. Yao, K. F. Ssu, and W. K. Fuchs. Message logging in mobile computing. Proceedings of IEEE Fault-Tolerant Computing Symposium, pages 294--301, June 1999.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC