Dr. Dobb's Journal - April 2008 - (Page 34) d04nels_p4ds 2/13/08 10:12 AM Page 34 Core Technology THE BYZANTINE GENERALS PROBLEM Figure 5: The tree after calculating the output values. sends those messages to all other processes, including itself. (The “where allowed” just means that the process skips any messages where adding its process ID to the list would create a cycle in the string of process IDs.) For example, suppose that in Round 0 that P1, a faulty general, told P2, P3, and P4 that the command value was 0, and told P5, P6, and P7 that the command value was 1. In Round 1, the messages in Table 1 would be sent. Things get more complicated in the second round. From the previous rule, we know that each process now has six values that it received in the previous round—one message from each of the three processes—and it needs to send each of those messages to all of the other processes, which might mean each process would send 36 messages out. In Table 1, I showed the messages being sent to all six processes, which is redundant because the same messages are broadcast to all processes. For Round 2, I just show the set of messages that each process sends to all of its neighbors. The six messages that P2 received in Round 1 were {0,12}, {0,13}, {0,14}, {1,15}, {1,16}, and {1,17}. According to the earlier definition, P2 will append its process ID to the path and forward each resulting message to all other processes. The possible messages it could broadcast in Round 2 are {0,122}, {0,132}, {0,142}, {1,152}, {1,162}, and {1,172}. The first message, {1,122}, contains a cycle in the path value of the tuple, so it is tossed out, leaving five messages to be sent to all processes. The first message that P2 is sending in Round 2, {0,132}, is equivalent to saying, “P2 is telling you that in round 1 P3 told it that in round 0 that P1 (the General) told it that the value was 0.” The five messages shown in P2’s column in Table 2 are sent to all six lieutenant processes, including itself. It’s easy to see that as the number of processes increases, the number of messages being exchanged starts to go up rapidly. If there are N processes, each process sends N–1 messages in Round 1, then (N–1)*(N–2) in Round 2, and (N–1)*(N–2)*(N–3) in Round 3. That can add up to a lot of messages in a big system. 34 Dr. Dobb’s Journal l www.ddj.com l April 2008 http://www.ddj.com
Table of Contents Feed for the Digital Edition of Dr. Dobb's Journal - April 2008 Dr. Dobb's Journal - April 2008 Contents Hmmmm Alia Vox Developer Diaries Dr. Dobb's Excellence in Programming Award Conversations Fast String Search on Multicore Processors The Byzantine Generals Problem Optimizing Math-Intensive Applications with Fixed-Point Arithmetic Random Numbers in a Range Using Generic Programming The Agile Edge Effective Concurrency Swaine's Flames Dr. Dobb's Journal - April 2008 Dr. Dobb's Journal - April 2008 - Dr. Dobb's Journal - April 2008 (Page Cover1) Dr. Dobb's Journal - April 2008 - Dr. Dobb's Journal - April 2008 (Page Cover2) Dr. Dobb's Journal - April 2008 - Dr. Dobb's Journal - April 2008 (Page 1) Dr. Dobb's Journal - April 2008 - Dr. Dobb's Journal - April 2008 (Page 2) Dr. Dobb's Journal - April 2008 - Dr. Dobb's Journal - April 2008 (Page 3) Dr. Dobb's Journal - April 2008 - Contents (Page 4) Dr. Dobb's Journal - April 2008 - Contents (Page 5) Dr. Dobb's Journal - April 2008 - Hmmmm (Page 6) Dr. Dobb's Journal - April 2008 - Hmmmm (Page 7) Dr. Dobb's Journal - April 2008 - Hmmmm (Page 8) Dr. Dobb's Journal - April 2008 - Hmmmm (Page 9) Dr. Dobb's Journal - April 2008 - Alia Vox (Page 10) Dr. Dobb's Journal - April 2008 - Alia Vox (Page 11) Dr. Dobb's Journal - April 2008 - Developer Diaries (Page 12) Dr. Dobb's Journal - April 2008 - Developer Diaries (Page 13) Dr. Dobb's Journal - April 2008 - Developer Diaries (Page 14) Dr. Dobb's Journal - April 2008 - Developer Diaries (Page 15) Dr. Dobb's Journal - April 2008 - Dr. Dobb's Excellence in Programming Award (Page 16) Dr. Dobb's Journal - April 2008 - Dr. Dobb's Excellence in Programming Award (Page 17) Dr. Dobb's Journal - April 2008 - Conversations (Page 18) Dr. Dobb's Journal - April 2008 - Conversations (Page 19) Dr. Dobb's Journal - April 2008 - Fast String Search on Multicore Processors (Page 20) Dr. Dobb's Journal - April 2008 - Fast String Search on Multicore Processors (Page 21) Dr. Dobb's Journal - April 2008 - Fast String Search on Multicore Processors (Page 22) Dr. Dobb's Journal - April 2008 - Fast String Search on Multicore Processors (Page 23) Dr. Dobb's Journal - April 2008 - Fast String Search on Multicore Processors (Page 24) Dr. Dobb's Journal - April 2008 - Fast String Search on Multicore Processors (Page 25) Dr. Dobb's Journal - April 2008 - Fast String Search on Multicore Processors (Page 26) Dr. Dobb's Journal - April 2008 - Fast String Search on Multicore Processors (Page 27) Dr. Dobb's Journal - April 2008 - Fast String Search on Multicore Processors (Page 28) Dr. Dobb's Journal - April 2008 - Fast String Search on Multicore Processors (Page 29) Dr. Dobb's Journal - April 2008 - The Byzantine Generals Problem (Page 30) Dr. Dobb's Journal - April 2008 - The Byzantine Generals Problem (Page 31) Dr. Dobb's Journal - April 2008 - The Byzantine Generals Problem (Page 32) Dr. Dobb's Journal - April 2008 - The Byzantine Generals Problem (Page 33) Dr. Dobb's Journal - April 2008 - The Byzantine Generals Problem (Page 34) Dr. Dobb's Journal - April 2008 - The Byzantine Generals Problem (Page 35) Dr. Dobb's Journal - April 2008 - The Byzantine Generals Problem (Page 36) Dr. Dobb's Journal - April 2008 - The Byzantine Generals Problem (Page 37) Dr. Dobb's Journal - April 2008 - Optimizing Math-Intensive Applications with Fixed-Point Arithmetic (Page 38) Dr. Dobb's Journal - April 2008 - Optimizing Math-Intensive Applications with Fixed-Point Arithmetic (Page 39) Dr. Dobb's Journal - April 2008 - Optimizing Math-Intensive Applications with Fixed-Point Arithmetic (Page 40) Dr. Dobb's Journal - April 2008 - Optimizing Math-Intensive Applications with Fixed-Point Arithmetic (Page 41) Dr. Dobb's Journal - April 2008 - Optimizing Math-Intensive Applications with Fixed-Point Arithmetic (Page 42) Dr. Dobb's Journal - April 2008 - Optimizing Math-Intensive Applications with Fixed-Point Arithmetic (Page 43) Dr. Dobb's Journal - April 2008 - Random Numbers in a Range Using Generic Programming (Page 44) Dr. Dobb's Journal - April 2008 - Random Numbers in a Range Using Generic Programming (Page 45) Dr. Dobb's Journal - April 2008 - Random Numbers in a Range Using Generic Programming (Page 46) Dr. Dobb's Journal - April 2008 - Random Numbers in a Range Using Generic Programming (Page 47) Dr. Dobb's Journal - April 2008 - Random Numbers in a Range Using Generic Programming (Page 48) Dr. Dobb's Journal - April 2008 - The Agile Edge (Page 49) Dr. Dobb's Journal - April 2008 - The Agile Edge (Page 50) Dr. Dobb's Journal - April 2008 - The Agile Edge (Page 51) Dr. Dobb's Journal - April 2008 - Effective Concurrency (Page 52) Dr. Dobb's Journal - April 2008 - Effective Concurrency (Page 53) Dr. Dobb's Journal - April 2008 - Effective Concurrency (Page 54) Dr. Dobb's Journal - April 2008 - Effective Concurrency (Page 55) Dr. Dobb's Journal - April 2008 - Swaine's Flames (Page 56) Dr. Dobb's Journal - April 2008 - Swaine's Flames (Page Cover3) Dr. Dobb's Journal - April 2008 - Swaine's Flames (Page Cover4)
For optimal viewing of this digital publication, please enable JavaScript and then refresh the page. If you would like to try to load the digital publication without using Flash Player detection, please click here.