README revision 98524
1@(#) $Header: /tcpdump/master/tcpdump/README,v 1.60 2002/01/21 11:16:33 mcr Exp $ (LBL) 2 3TCPDUMP 3.7.1 4Now maintained by "The Tcpdump Group" 5See www.tcpdump.org 6 7Please send inquiries/comments/reports to tcpdump-workers@tcpdump.org 8 9Anonymous CVS is available via: 10 cvs -d cvs.tcpdump.org:/tcpdump/master login 11 (password "anoncvs") 12 cvs -d cvs.tcpdump.org:/tcpdump/master checkout tcpdump 13 14Version 3.7.1 of TCPDUMP can be retrived with the CVS tag "tcpdump_3_7rel1": 15 cvs -d cvs.tcpdump.org:/tcpdump/master checkout -r tcpdump_3_7rel1 tcpdump 16 17Please send patches against the master copy to patches@tcpdump.org. 18 19formerly from Lawrence Berkeley National Laboratory 20 Network Research Group <tcpdump@ee.lbl.gov> 21 ftp://ftp.ee.lbl.gov/tcpdump.tar.Z (3.4) 22 23This directory contains source code for tcpdump, a tool for network 24monitoring and data acquisition. This software was originally 25developed by the Network Research Group at the Lawrence Berkeley 26National Laboratory. The original distribution is available via 27anonymous ftp to ftp.ee.lbl.gov, in tcpdump.tar.Z. More recent 28development is performed at tcpdump.org, http://www.tcpdump.org/ 29 30Tcpdump uses libpcap, a system-independent interface for user-level 31packet capture. Before building tcpdump, you must first retrieve and 32build libpcap, also originally from LBL and now being maintained by 33tcpdump.org; see http://www.tcpdump.org/ . 34 35Once libpcap is built (either install it or make sure it's in 36../libpcap), you can build tcpdump using the procedure in the INSTALL 37file. 38 39The program is loosely based on SMI's "etherfind" although none of the 40etherfind code remains. It was originally written by Van Jacobson as 41part of an ongoing research project to investigate and improve tcp and 42internet gateway performance. The parts of the program originally 43taken from Sun's etherfind were later re-written by Steven McCanne of 44LBL. To insure that there would be no vestige of proprietary code in 45tcpdump, Steve wrote these pieces from the specification given by the 46manual entry, with no access to the source of tcpdump or etherfind. 47 48Over the past few years, tcpdump has been steadily improved by the 49excellent contributions from the Internet community (just browse 50through the CHANGES file). We are grateful for all the input. 51 52Richard Stevens gives an excellent treatment of the Internet protocols 53in his book ``TCP/IP Illustrated, Volume 1''. If you want to learn more 54about tcpdump and how to interpret its output, pick up this book. 55 56Some tools for viewing and analyzing tcpdump trace files are available 57from the Internet Traffic Archive: 58 59 http://www.acm.org/sigcomm/ITA/ 60 61Another tool that tcpdump users might find useful is tcpslice: 62 63 ftp://ftp.ee.lbl.gov/tcpslice.tar.Z 64 65It is a program that can be used to extract portions of tcpdump binary 66trace files. See the above distribution for further details and 67documentation. 68 69Problems, bugs, questions, desirable enhancements, etc. 70should be sent to the address "tcpdump-workers@tcpdump.org". 71 72Source code contributions, etc. should be sent to the email address 73"patches@tcpdump.org". 74 75Current versions can be found at www.tcpdump.org 76 77 - The TCPdump team 78 79original text by: Steve McCanne, Craig Leres, Van Jacobson 80 81------------------------------------- 82This directory also contains some short awk programs intended as 83examples of ways to reduce tcpdump data when you're tracking 84particular network problems: 85 86send-ack.awk 87 Simplifies the tcpdump trace for an ftp (or other unidirectional 88 tcp transfer). Since we assume that one host only sends and 89 the other only acks, all address information is left off and 90 we just note if the packet is a "send" or an "ack". 91 92 There is one output line per line of the original trace. 93 Field 1 is the packet time in decimal seconds, relative 94 to the start of the conversation. Field 2 is delta-time 95 from last packet. Field 3 is packet type/direction. 96 "Send" means data going from sender to receiver, "ack" 97 means an ack going from the receiver to the sender. A 98 preceding "*" indicates that the data is a retransmission. 99 A preceding "-" indicates a hole in the sequence space 100 (i.e., missing packet(s)), a "#" means an odd-size (not max 101 seg size) packet. Field 4 has the packet flags 102 (same format as raw trace). Field 5 is the sequence 103 number (start seq. num for sender, next expected seq number 104 for acks). The number in parens following an ack is 105 the delta-time from the first send of the packet to the 106 ack. A number in parens following a send is the 107 delta-time from the first send of the packet to the 108 current send (on duplicate packets only). Duplicate 109 sends or acks have a number in square brackets showing 110 the number of duplicates so far. 111 112 Here is a short sample from near the start of an ftp: 113 3.00 0.20 send . 512 114 3.20 0.20 ack . 1024 (0.20) 115 3.20 0.00 send P 1024 116 3.40 0.20 ack . 1536 (0.20) 117 3.80 0.40 * send . 0 (3.80) [2] 118 3.82 0.02 * ack . 1536 (0.62) [2] 119 Three seconds into the conversation, bytes 512 through 1023 120 were sent. 200ms later they were acked. Shortly thereafter 121 bytes 1024-1535 were sent and again acked after 200ms. 122 Then, for no apparent reason, 0-511 is retransmitted, 3.8 123 seconds after its initial send (the round trip time for this 124 ftp was 1sec, +-500ms). Since the receiver is expecting 125 1536, 1536 is re-acked when 0 arrives. 126 127packetdat.awk 128 Computes chunk summary data for an ftp (or similar 129 unidirectional tcp transfer). [A "chunk" refers to 130 a chunk of the sequence space -- essentially the packet 131 sequence number divided by the max segment size.] 132 133 A summary line is printed showing the number of chunks, 134 the number of packets it took to send that many chunks 135 (if there are no lost or duplicated packets, the number 136 of packets should equal the number of chunks) and the 137 number of acks. 138 139 Following the summary line is one line of information 140 per chunk. The line contains eight fields: 141 1 - the chunk number 142 2 - the start sequence number for this chunk 143 3 - time of first send 144 4 - time of last send 145 5 - time of first ack 146 6 - time of last ack 147 7 - number of times chunk was sent 148 8 - number of times chunk was acked 149 (all times are in decimal seconds, relative to the start 150 of the conversation.) 151 152 As an example, here is the first part of the output for 153 an ftp trace: 154 155 # 134 chunks. 536 packets sent. 508 acks. 156 1 1 0.00 5.80 0.20 0.20 4 1 157 2 513 0.28 6.20 0.40 0.40 4 1 158 3 1025 1.16 6.32 1.20 1.20 4 1 159 4 1561 1.86 15.00 2.00 2.00 6 1 160 5 2049 2.16 15.44 2.20 2.20 5 1 161 6 2585 2.64 16.44 2.80 2.80 5 1 162 7 3073 3.00 16.66 3.20 3.20 4 1 163 8 3609 3.20 17.24 3.40 5.82 4 11 164 9 4097 6.02 6.58 6.20 6.80 2 5 165 166 This says that 134 chunks were transferred (about 70K 167 since the average packet size was 512 bytes). It took 168 536 packets to transfer the data (i.e., on the average 169 each chunk was transmitted four times). Looking at, 170 say, chunk 4, we see it represents the 512 bytes of 171 sequence space from 1561 to 2048. It was first sent 172 1.86 seconds into the conversation. It was last 173 sent 15 seconds into the conversation and was sent 174 a total of 6 times (i.e., it was retransmitted every 175 2 seconds on the average). It was acked once, 140ms 176 after it first arrived. 177 178stime.awk 179atime.awk 180 Output one line per send or ack, respectively, in the form 181 <time> <seq. number> 182 where <time> is the time in seconds since the start of the 183 transfer and <seq. number> is the sequence number being sent 184 or acked. I typically plot this data looking for suspicious 185 patterns. 186 187 188The problem I was looking at was the bulk-data-transfer 189throughput of medium delay network paths (1-6 sec. round trip 190time) under typical DARPA Internet conditions. The trace of the 191ftp transfer of a large file was used as the raw data source. 192The method was: 193 194 - On a local host (but not the Sun running tcpdump), connect to 195 the remote ftp. 196 197 - On the monitor Sun, start the trace going. E.g., 198 tcpdump host local-host and remote-host and port ftp-data >tracefile 199 200 - On local, do either a get or put of a large file (~500KB), 201 preferably to the null device (to minimize effects like 202 closing the receive window while waiting for a disk write). 203 204 - When transfer is finished, stop tcpdump. Use awk to make up 205 two files of summary data (maxsize is the maximum packet size, 206 tracedata is the file of tcpdump tracedata): 207 awk -f send-ack.awk packetsize=avgsize tracedata >sa 208 awk -f packetdat.awk packetsize=avgsize tracedata >pd 209 210 - While the summary data files are printing, take a look at 211 how the transfer behaved: 212 awk -f stime.awk tracedata | xgraph 213 (90% of what you learn seems to happen in this step). 214 215 - Do all of the above steps several times, both directions, 216 at different times of day, with different protocol 217 implementations on the other end. 218 219 - Using one of the Unix data analysis packages (in my case, 220 S and Gary Perlman's Unix|Stat), spend a few months staring 221 at the data. 222 223 - Change something in the local protocol implementation and 224 redo the steps above. 225 226 - Once a week, tell your funding agent that you're discovering 227 wonderful things and you'll write up that research report 228 "real soon now". 229