ATM-based cluster computing for multi-problem domains
This study evaluates the performance of an Asynchronous Transfer Mode (ATM) local area network (LAN) for general as well as parallel distributed computing. General distributed computing uses client-server based applications that employ Remote Procedure Call (RPC) on top of the TCP/UDP/IP protocol. These applications typically require high throughput, good response time, and fairness. In contrast, parallel applications favor much simpler models of computation which require more direct access to data among processors. To efficiently run these programs, the network needs to achieve hardware speed. This paper describe our experience in building a multi-programmed distributed computing environment using Digital Equipment Corporation`s (DEC) AN2 ATM switch and Alpha workstations. We extend our study to include more elaborate network using simulation results.
- Research Organization:
- Sandia National Lab. (SNL-NM), Albuquerque, NM (United States)
- Sponsoring Organization:
- USDOE Office of Energy Research, Washington, DC (United States)
- DOE Contract Number:
- AC04-94AL85000
- OSTI ID:
- 415338
- Report Number(s):
- SAND-96-8249; ON: DE97050469; TRN: 97:000101
- Resource Relation:
- Other Information: PBD: Aug 1996
- Country of Publication:
- United States
- Language:
- English
Similar Records
Distributed network computing over local ATM networks
Congestion control over ATM local area networks