Cluster I/O with River
- 1 May 1999
- proceedings article
- Published by Association for Computing Machinery (ACM)
Abstract
We introduce River, a data-flow programming environment and I/O substrate for clusters of computers. River is designed to provide max- imum performance in the common case — even in the face of non- uniformities in hardware, software, and workload. River is based on two simple design features: a high-performance distributed queue, and a storage redundancy mechanism called graduated declustering. We have implemented a number of data-intensive applications on River, which validate our design with near-ideal performance in a variety of non-uniform performance scenarios.Keywords
This publication has 25 references indexed in Scilit:
- The Vesta parallel file systemACM Transactions on Computer Systems, 1996
- Myrinet: a gigabit-per-second local area networkIEEE Micro, 1995
- Loading databases using dataflow parallelismACM SIGMOD Record, 1994
- Parallel database systemsCommunications of the ACM, 1992
- Active messagesACM SIGARCH Computer Architecture News, 1992
- Orca: a language for parallel programming of distributed systemsIEEE Transactions on Software Engineering, 1992
- Encapsulation of parallelism in the Volcano query processing systemACM SIGMOD Record, 1990
- A performance analysis of the gamma database machineACM SIGMOD Record, 1988
- Join processing in database systems with large main memoriesACM Transactions on Database Systems, 1986
- End-to-end arguments in system designACM Transactions on Computer Systems, 1984