The increasing need for data sharing in large-scale distributed systems may place a heavy burden on critical resources such as file servers and networks. Our examination of the workload in one large commercial engineering environment shows that wide-spread sharing of unstable files among tens to hundreds of users is common. Traditional clientbased file cacheing techniques are not scalable in such environments. We propose Frolic, a scheme for cluster-based file replication in large-scale distributed file systems. A cluster is a group of workstations and one or more file servers on a local area network. Large distributed systems may have tens or hundreds of clusters connected by a backbone network. By dynamically creating and maintaining replicas of shared files on the file servers in the clusters using those files, we effectively reduce reliance on central servers supporting such files, as well as reduce the distances between the accessing sites and data. We propose and study algorithms...

A common way to address scalability requirements of distributed services is to employ server replication and client caching of objects that encapsulate the service state. The performance of such a system could depend very much on the protocol implemented by the system to maintain consistency among object copies. We explore scalable consistency protocols that never require synchronization and communication between all nodes that have copies of related objects. We achieve this by developing a novel approach called local consistency (LC). LC based protocols can provide increased flexibility and efficiency by allowing nodes control over how and when they become aware of updates to cached objects. We develop two protocols for implementing strong consistency using this approach and demonstrate that they scale better than a traditional invalidation based consistency protocol along the system load and geographic distribution dimensions of scale.

This paper describes the design, implementation and performance of a family of protocols for survivable, decentralized data storage. These protocols exploit storage-node versioning to efficiently achieve strong consistency semantics. These protocols allow erasure-codes to be used that achieve network and storage efficiency (and optionally data confidentiality in the face of server compromise). The protocol family is general in that its parameters accommodate a wide range of fault and timing assumptions, up to asynchrony and Byzantine faults of both storage-nodes and clients, with no changes to server implementation or client-server interface. Measurements of a prototype storage system using these protocols show that the protocol performs well under various system model assumptions, numbers of failures tolerated, and degrees of reader-writer concurrency.

This paper describes a consistency protocol that exploits versioning storage-nodes. The protocol provides linearizability with the possibility of read aborts in an asynchronous system that may suffer client and storage-node crash failures. The protocol supports both replication and erasure coding (which precludes post hoc repair of partial-writes), and avoids the excess work of two-phase commits. Versioning storagenodes allow the protocol to avoid excess communication in the common case of no write sharing and no failures of writing clients.

To Candi, for her love, support, and many years of patience. ACKNOWLEDGEMENTS Those who praise “thinking outside the box ” have never had to deal with a gradu-ate student who chose a thesis topic outside every one of his department’s “boxes.” Many people have gone above and beyond the call of duty to provide me the re-sources and advice needed to finish this dissertation. First, I want to thank Kimberly Keeton, Ralph Becker-Szendy, John Wilkes, Allistar Vietch, and everybody else in the Storage Systems Department at HP Labs for granting me the privilege to join them for a summer, and for continuing to support my research after I had returned to Georgia Tech. Second, I want to thank Ann Chervenak, Ken Mackenzie, Kishore Ramachandran, and all the other faculty at Georgia Tech for providing the moral and financial support that allowed me to pursue my storage systems research. Third, I thank Chad Huneycutt, Josh Fryman, and all the other students at Georgia Tech who spent time editing my papers and critiquing my talks. Nobody was under any

Distributed File Systems have long utilized file caching techniques to improve performance. In many DFS’s clients are allowed to update the cached replicas of files, necessitating a variety of mechanisms that ensure the consistency of other replicas of these files across the network. This problem becomes complicated if there are many unstable files on the network, and especially so if there are no central servers. Surprisingly, there has not been much research into the access patterns of shared files. While researchers have found that there are relatively few unstable files, this claim has been disputed by others, and the last such study that I am aware of dates back to 1992. Even more strikingly, I am aware of no detailed studies of file reading and writing patterns at all. In an attempt to fill this gap, I analyzed file access patterns in Coda Distributed File System using traces collected at Carnegie Mellon University over a period of approximately two years. I found that 1) Most files analyzed are stable; 2) Most unstable files tend to be updated and read by only one computer, but the computer that reads a file tends to be different from the one that updates it; and 3) A vast majority of files are read by the same computer that created them. 1.

For many distributed data intensive applications, the default remote invocation of CORBA objects by clients is not acceptable because of performance degradation. Caching enables clients to invoke operations locally on distributed objects instead of fetching them from remote servers. This paper addresses the design and implementation of a specific caching approach for CORBA-based systems. We propose a new removal algorithm which uses a double linked structure and a hash table for eviction. We also present a new variation of optimistic two phase locking for consistency control, which does not require a lock at the client side by using a perprocess caching design. With the experiments we have performed, we demonstrate that the proposed caching approach provides an important performance gain: the caching with half buffer saves up to 45% of access time and the caching with full buffer saves up to 50% of access time. The Common Object Request Broker Architecture (CORBA) provides several adv...