
































Study with the several resources on Docsity
Earn points by helping other students or get them with a premium plan
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
Community
Ask the community for help and clear up your study doubts
Discover the best universities in your country according to Docsity users
Free resources
Download our free guides on studying techniques, anxiety management strategies, and thesis advice from Docsity tutors
Material Type: Notes; Class: Distributed Computer Systems; Subject: Computer Science; University: Brown University; Term: Spring 2008;
Typology: Study notes
1 / 40
This page cannot be seen from the preview
Don't miss anything!
This lecture includes material from Tanenbaum and Van Steen as well as from Ugur Cetintemel.Cetintemel.
A good exposition on MPI (message-passing interface) can be found at https://computing.llnl.gov/tutorials/mpi/. A discussion of performance issues can be foundhttps://computing.llnl.gov/tutorials/mpi/. A discussion of performance issues can be found at https://computing.llnl.gov/tutorials/mpi_performance/.
In the MPI model, senders provide, in application buffers (program variables), data to be sent. The act of sending (calling one of the send routines) tells the system that the data issent. The act of sending (calling one of the send routines) tells the system that the data is ready to go. At this point the data is transferred to system buffer on the sender’s side, and then from there to a system buffer on the receiver’s side. Receivers must call a receive routine to have data transferred from the receive-side system buffer to a receive-side application buffer (the receive call may be executed before data arrives at the system buffer, even before send is called on the system side).
The diagram is from https://computing.llnl.gov/tutorials/mpi/.
The diagram is from https://computing.llnl.gov/tutorials/mpi/.
The diagram is from https://computing.llnl.gov/tutorials/mpi/.
The diagram is from https://computing.llnl.gov/tutorials/mpi/.
This is Figure 4-18 from Tanenbaum and Van Steen: Basic interface to a queue in a message-queuing system.message-queuing system.
This is Figure 4-19 from Tanenbaum and Van Steen. The relationship between queue-level addressing and network-level addressing.addressing and network-level addressing.
This is Figure 4-21 from Tanenbaum and Van Steen: The general organization of a message broker in a message-queuing system.broker in a message-queuing system.
This is Figure 4-22 from Tanenbaum and Van Steen: General organization of IBM’s message-queuing system. “MCA” stands formessage-queuing system. “MCA” stands for messagemessage channelchannel agentagent : there’s one on each end: there’s one on each end of a communication channel. A paper describing what Websphere is used for can be found at ftp://ftp.software.ibm.com/software/integration/wmq/WS_MQ_Messaging_Backbone_for_SO A.pdf (caution: this is a marketing paper!).