Dan Duchamp

An analytical approach to file prefetching

File prefetching is an effective technique for improving file access performance. In this paper, we present a file prefetching mechanism that is based on on-line analytic modeling of interesting system events and is transparent to higher levels. The mechanism, incorporated into a clients file cache manager, seeks to build semantic structures that capture the intrinsic correlations between file accesses. It then heuristically uses these structures to represent distinct file usage patterns and exploits them to prefetch files from a file server. We show results of a simulation study and of a working implementation. Measurements suggest that our method can predict future file accesses with an accuracy around 90%, that it can reduce cache miss rate by up to 47% and application latency by up to 40%. Our method imposes little overhead, even under antagonistic circumstances.

Detection and exploitation of file working sets

The work habits of many individuals yield file access patterns that are quite pronounced and can be regarded as defining working sets of files used for particular applications. A client management technique for detecting these patterns and then exploiting them to successfully prefetch files from servers is described. Trace-driven simulations show the technique substantially increases file cache hit rate in a single-user environment. Successful file prefetching carries three major advantages: applications run faster, there is less burst load placed on the network, and properly loaded client caches can better survive network outages. The technique requires little extra code, and-because it is simply an augmentation of the standard LRU client cache management algorithm-is easily incorporated into existing software.<<ETX>>

Issues in wireless mobile computing

The trend toward possible computers with interfaces to wireless network technologies is examined. The question of needed research and specification of hardware is discussed, the major challenges, such as services for the mobile user, mobile internetworking and adjusting to new technologies are described, and a listing of outstanding questions which must be resolved is presented.<<ETX>>

Measured performance of a wireless LAN

The performance of a high-speed commercial spread-spectrum wireless LAN that uses the CSMA/CA multiple-access strategy was studied. Using synthetic workloads, packet capture success rather than signal propagation characteristics was measured. Specifically throughput, packet loss rates, range, and patterns of errors within packets were measured. It is concluded that CSMA/CA is quite successful in allocating bandwidth under stress, but that packet capture rate degrades very quickly once the LANs effective range is exceeded. Hence, network maintainers should plan the layout of wireless networks at least as carefully as they plan wired networks.<<ETX>>

Intelligent communication filtering for limited bandwidth environments

We advocate a new architecture for distributed systems supporting mobility. In this architecture an intermediary is interposed between client and server. Its purpose is to filter or delay all but the most essential data that would travel over the slow link to the mobile host, as instructed by the client. The goal of our work is to provide a general framework to ease the installation and management of application specific filtering mechanisms.

Service interface and replica management algorithm for mobile file system clients

Portable computers are now common, a fact that raises the possibility that file service clients might move on a regular basis. This new development requires rethinking some features of distributed file system design. The authors argue that existing approaches to file replica management would not cope well with the likely behavior of mobile clients, and they present their solution; a lazy server-based update operation. This operation facilitates fast, scalable, and highly fault tolerant implementations of both read and write operations in the usual case. To cope with the weak semantics of the update operation, the authors propose a new file system service interface that allows applications to opt for UNIX semantics by use of a slower, less fault-tolerant read operation.<<ETX>>

Analysis of transaction management performance

There is currently much interest in incorporating transactions into both operating systems and general-purpose programming languages. This paper provides a detailed examination of the design and performance of the transaction manager of the Camelot system. Camelot is a transaction facility that provides a rich model of transactions intended to support a wide variety of general-purpose applications. The transaction managers principal function is to execute the protocols that ensure atomicity. The conclusions of this study are: a simple optimization to two-phase commit reduces logging activity of distributed transactions; non-blocking commit is practical for some applications; multithreaded design improves throughput provided that log batching is used; multi-casting reduces the variance of distributed commit protocols in a LAN environment; and the performance of transaction mechanisms such as Camelot depend heavily upon kernel performance.

General purpose proxies: solved and unsolved problems

Proxies are becoming increasingly common. One use of a proxy is to address network heterogeneity, which can arise especially in mobile computing. In this setting, a proxy can process the traffic flowing to and from a network limited mobile host, damping the variations in application performance as well as providing other benefits such as reduced cost and increased security. We describe the issues we faced in designing such a proxy system, how we solved some problems, and why others could not be solved.

An interface to support lazy replicated file service

The authors argue that, if they are going to assume low-sharing workloads, then it is possible to build more efficient implementations that provide the same (or stronger) semantics by making fullest use of the assumption. Accordingly, they take an extreme point on the spreading/searching tradeoff: the searcher is made responsible for all the work. Thus, spreading can be made asynchronous and hence very fast from the callers viewpoint. They employ an important optimization to eliminate most of the searching work when there is little sharing. Their technique carries other advantages as well: simple and flexible recovery algorithms, and suitability for mobile computing.<<ETX>>