CSC/ECE 506 Spring 2012/2a bm: Difference between revisions
No edit summary |
No edit summary |
||
| Line 3: | Line 3: | ||
=== Background === | === Background === | ||
Early distributed computer systems relied almost exclusively on '''message passing'''(MP) to communicate with one another, and this technique is still widely used today. In a message passing model, each | Early distributed computer systems relied almost exclusively on '''message passing'''(MP) to communicate with one another, and this technique is still widely used today. In a message passing model, each processor's local memory is isolated from that of the rest of the system. Processes or objects can send or receive messages in order to communicate and this process can occur in a synchronous or asynchronous manner. In distributed systems, and particularly for certain types of programs, the message passing model can become overly burdensome to the programmer as tracking data movement and maintaining data integrity can become quite challenging with many control threads. A shared address or shared-memory system, however, can provide a programming model that simplifies data sharing via uniform mechanisms of data structure reads and writes on common memory. Current distributed systems seek to take advantage both SAS and MP programming model principles in hybrid systems. | ||
=== Distributed Shared Memory (DSM) === | === Distributed Shared Memory (DSM) === | ||
Revision as of 22:52, 26 January 2012
SAS programming on distributed-memory machines
Shared Address Space (SAS) programming on distributed memory machines is a programming abstraction that provides less development effort than that of the traditional method of Message Passing (MP) on distributed memory machines, such as clusters of servers. Distributed systems are groups of computers that communicate through a network and share a common work goal. Distributed systems typically do not physically share the same memory (tightly coupled) but rather each processor or group of processors (e.g. Symmetric Multiprocessor (SMP)) must depend on other mechanisms besides direct memory access in order to communicate. Relevant concepts include memory coherence, types of memory access, data and process synchronization, and performance.
Background
Early distributed computer systems relied almost exclusively on message passing(MP) to communicate with one another, and this technique is still widely used today. In a message passing model, each processor's local memory is isolated from that of the rest of the system. Processes or objects can send or receive messages in order to communicate and this process can occur in a synchronous or asynchronous manner. In distributed systems, and particularly for certain types of programs, the message passing model can become overly burdensome to the programmer as tracking data movement and maintaining data integrity can become quite challenging with many control threads. A shared address or shared-memory system, however, can provide a programming model that simplifies data sharing via uniform mechanisms of data structure reads and writes on common memory. Current distributed systems seek to take advantage both SAS and MP programming model principles in hybrid systems.
Implementations
There are
Hardware
Algorithms
Evolution
References
- Shan, H.; Singh, J.P.; Oliker, L.; Biswas, R.; , "Message passing vs. shared address space on a cluster of SMPs," Parallel and Distributed Processing Symposium., Proceedings 15th International , vol., no., pp.8 pp., Apr 2001
- Protic, J.; Tomasevic, M.; Milutinovic, V.; , "Distributed shared memory: concepts and systems," Parallel & Distributed Technology: Systems & Applications, IEEE , vol.4, no.2, pp.63-71, Summer 1996
- Chandola, V. , "Design Issues in Implementation of Distributed Shared Memory in User Space,"