The tradeoffs between consistency, performance, and availability are well understood. Traditionally, however, designers of replicated systems have been forced to choose from either strong consistency guarantees or none at all. This paper explores the semantic space between traditional strong and optimistic consistency models for replicated services. We argue that an important class of applications can tolerate relaxed consistency, but benefit from bounding the maximum rate of inconsistent access in an application-specific manner. Thus, we develop a set of metrics, Numerical Error, Order Error, and Staleness, to capture the consistency spectrum. We then present the design and implementation of TACT, a middleware layer that enforces arbitrary consistency bounds among replicas using these metrics. Finally, we show that three replicated applications demonstrate significant semantic and performance benefits from using our framework. 1
|
749
|
Disconnected Operation in the CODA file system
– Satyanarayanan, Kistler
- 1992
|
|
592
|
the Ordering of Events in a Distributed System
– Time
- 1978
|
|
282
|
Managing update conflicts in bayou, a weakly connected replicated storage system
– Terry, Theimer, et al.
- 1995
|
|
240
|
Locality-Aware Request Distribution in Cluster-Based Network Servers
– Pai, Aron, et al.
- 1998
|
|
207
|
Flexible update propagation for weakly consistent replication
– Petersen, Spreitzer, et al.
- 1997
|
|
195
|
Data caching issues in an information retrieval system
– Alonso, Barbará, et al.
- 1990
|
|
148
|
Replica control in distributed system: An asynchronous approach
– Pu, Leff
- 1991
|
|
123
|
Synchronizing a Database to Improve Freshness
– Cho, García-Molina
- 2000
|
|
102
|
Implementation of the ficus replicated file system
– Guy, Heidemann, et al.
- 1990
|
|
101
|
Weak Consistency group communications and memberships
– Golding
- 1992
|
|
99
|
Efficient optimistic concurrency control using loosely synchronized clocks
– Adya, Gruber, et al.
|
|
96
|
Manageability, availability and performance in Porcupine: A highly scalable, cluster-based mail service
– Saito, Bershad, et al.
- 1999
|
|
60
|
Decentralized replicated-object protocols
– Keleher
- 1999
|
|
48
|
yield and scalable tolerant systems
– Harvest
- 1999
|
|
48
|
Ssp: a semantics-based protocol for real-time data access
– Kuo, Mok
- 1993
|
|
43
|
Application semantics and concurrency control of real-time dataintensive applications
– Kuo, Mok
- 1992
|
|
42
|
Rumor: Mobile data access through optimistic peer-to-peer replication
– Guy, Reiher, et al.
- 1998
|
|
38
|
Designing and implementing asynchronous collaborative applications with Bayou
– Edwards, Mynatt, et al.
- 1997
|
|
37
|
Consistency and orderability: semantics-based correctness criteria for databases
– Agrawal, Abbadi, et al.
- 1993
|
|
37
|
Temporal notions of synchronization and consistency in Beehive
– Singla, Ramachandran, et al.
- 1997
|
|
35
|
Efficient Numerical Error Bounding for Replicated Network Services
– Yu, Vahdat
- 2000
|
|
34
|
Providing Availability Using Lazy Replication
– Ladin, Liskov, et al.
- 1992
|
|
32
|
Bounded ignorance: A technique for increasing concurrency in replicated systems
– KRISHNAKUMAR, BERNSTEIN
- 1994
|
|
29
|
Tolerating bounded inconsistency for increasing concurrency in database systems
– Wong, Agrawal
- 1992
|
|
28
|
Limitations on database availability when networks partition
– Coan, Oki, et al.
- 1986
|
|
26
|
Timed consistency for shared distributed objects
– TORRES-ROJAS, AHAMAD, et al.
- 1999
|
|
20
|
Abbadi. Epidemic Quorums for Managing Replicated Data
– Holliday, Steinke, et al.
- 2000
|
|
16
|
Information Storage in a Decentralized Computer System
– Gifford
- 1983
|
|
15
|
A case for fluid replication
– NOBLE, FLEIS, et al.
- 2000
|
|
11
|
The Failure and Recovery Problem for Replicated Distributed Databases
– Bernstein, Goodman
- 1984
|
|
11
|
Distributed divergence control for epsilon serializability
– PU, HSEUSH, et al.
- 1993
|
|
11
|
Building Replicated Internet Services using TACT: A Toolkit for Tunable Availability and Consistency Tradeoffs
– Yu, Vahdat
- 2000
|
|
7
|
Improving Data Freshness in Lazy Master Schemes
– Pacitti, Simon, et al.
- 1998
|
|
6
|
The Proecss Group Appraoch to Reliable Distributed Computing
– Birman
- 1993
|
|
6
|
Bounded Aggregation: Offering a Precision-Performance Tradeoff in Replicated Systems
– Olston, Widom
- 1999
|
|
5
|
Session Guarantees for Weekly Consistent Replicated Data
– Terry, Demers, et al.
- 1994
|
