InfoMagic Internet Tools 1993 July

home *** CD-ROM | disk | FTP | other *** search

/ InfoMagic Internet Tools 1993 July / Internet Tools.iso / RockRidge / ip / trace / tcpdump-2.2.1 / README < prev next >

Wrap

Text File | 1992-05-26 | 13.5 KB | 332 lines

Fri May 22 17:19:41 PDT 1992 Version 2.2 This directory contains tcpdump version 2.2, the Berkely Packet Filter (BPF), and tcpslice, a tool for manipulating raw packet traces. Note that there was no 2.1 release. Version 2.1beta was released with the BSD Networking 2 tape, but we never got around to a general 2.1 release. Tcpdump runs on following platforms: machine os packet filter ------- -- ------------- hp300 4.3BSD Tahoe/Reno bpf sparc SunOS 4.x bpf, nit sun3 SunOS 3.5, SunOS 4.x bpf, nit Decstation Ultrix 4.0 (and higher) packetfilter IBM RT 4.3BSD enet 386/486 4.3BSD netII bpf BPF can be installed in SunOS kernels, provided you have source, 4.3BSD kernels, and is now standard in BSD systems. (The Networking II release from CSRG has BPF support. Tcpdump can be temporarily found in /usr/src/contrib.) See bpf/README for further details and an installation procedure. In addition to bug fixes, changes from version 2.0 include: - Easy access to icmp packets, via the 'icmp' keyword. For example, % tcpdump 'icmp[0] != 8 and icmp[0] != 0' matches non-echo/reply ICMP packets. - An improved filter code optimizer. - A multicast keyword. Also, the broadcast keyword can now be qualified with a protocol layer. For instance, "ip broadcast" and "ether multicast" are valid filters. - Support for monitoring the loopback interface (i.e. 'tcpdump -i lo'). Jeffrey Honig (jch@MITCHELL.CIT.CORNELL.EDU) contributed the kernel patches to netinet/if_loop.c. - Support for the Ungermann-Bass Ethernet on IBM/PC-RTs running AOS. Contact Jeffrey Honig for the diffs. - Decoding of EGP and OSPF packets, thanks to Jeffrey Honig. - The tcpslice program, thanks to Vern Paxson (vern@ee.lbl.gov). Check out the man page (tcpslice.1) for more information. The BPF kernel interface has changed and is not backward compatible with the interface from the 2.0 release. tcpdump-2.1 won't work with old versions of BPF, but will work with the CSRG Networking II release. If you've made the bpf changes to network drivers, you'll need to update them (see bpf/README). Also, if you've written any BPF applications, they may need some minor changes (with respect to ioctls). The BPF man page is improved and contains an explanation of the filtering language. Tcpdump's makefile has continued to evolve. Multiple platforms are supported using subdirectories. See INSTALL for more details. Problems, bugs, questions, desirable enhancements, etc., should be sent to the email address "tcpdump@ee.lbl.gov". - Steve McCanne (mccanne@ee.lbl.gov) Craig Leres (leres@ee.lbl.gov) Van Jacobson (van@ee.lbl.gov) ------------------------------ Old news: - A packet dumper has been added (thanks to Jeff Mogul of DECWRL). With this option, you can create an architecture independent binary trace file in real time, without the overhead of the packet printer. At a later time, the packets can be filtered (again) and printed. - BSD is supported. You must have BPF in your kernel. Since the filtering is now done in the kernel, fewer packets are dropped. In fact, with BPF and the packet dumper option, a measly Sun 3/50 can keep up with a busy network. - Compressed SLIP packets can now be dumped, provided you use our SLIP software and BPF. These packets are dumped as any other IP packet; the compressed headers are dumped with the '-e' option. - Machines with little-endian byte ordering are supported (thanks to Jeff Mogul). - Ultrix 4.0 is supported (also thanks to Jeff Mogul). - IBM RT and Stanford Enetfilter support has been added by Rayan Zachariassen <rayan@canet.ca>. Tcpdump has been tested under both the vanilla enetfilter interface, and the extended interface (#ifdef'd by IBMRTPC) present in the MERIT version of the enetfilter. - TFTP packets are now printed (requests only). - BOOTP packets are now printed. - SNMP packets are now printed. (thanks to John LoVerso of Xylogics). - Sparc architectures, including the Sparcstation-1, are now supported thanks to Steve McCanne and Craig Leres. - SunOS 4 is now supported thanks to Micky Liu of Columbia University (micky@cunixc.cc.columbia.edu). - IP options are now printed. - RIP packets are now printed. - There's a -v flag that prints out more information than the default (e.g., it will enable printing of IP ttl, tos and id) and -q flag that prints out less (e.g., it will disable interpretation of AppleTalk-in-UDP). - The grammar has undergone substantial changes (if you have an earlier version of tcpdump, you should re-read the manual entry). The most useful change is the addition of an expression syntax that lets you filter on arbitrary fields or values in the packet. E.g., "ip[0] > 0x45" would print only packets with IP options, "tcp[13] & 3 != 0" would print only TCP SYN and FIN packets. The most painful change is that concatenation no longer means "and" -- e.g., you have to say "host foo and port bar" instead of "host foo port bar". The up side to this down is that repeated qualifiers can be omitted, making most filter expressions shorter. E.g., you can now say "ip host foo and (bar or baz)" to look at ip traffic between hosts foo and bar or between hosts foo and baz. [The old way of saying this was "ip host foo and (ip host bar or ip host baz)".] ------------------------------ The program is loosely based on SMI's "etherfind" although none of the etherfind code remains. It was originally written by Van Jacobson, Lawrence Berkeley Laboratory, as part of an ongoing research project to investigate and improve tcp and internet gateway performance. The parts of the program originally taken from Sun's etherfind were later re-written by Steve McCanne of LBL. To insure that would be no vestige of proprietary code in tcpdump, Steve wrote these pieces from the specification given by the manual entry, with no access to the source of tcpdump or etherfind. The current versions of these files are available via anonymous ftp from host ftp.ee.lbl.gov (currently at address 128.3.254.68) file tcpdump.tar.Z (a compressed Unix tar file). This program is subject to the 'standard' Berkeley network software copyright: Copyright (c) 1988-1990 The Regents of the University of California. All rights reserved. Redistribution and use in source and binary forms, with or without modification, are permitted provided that: (1) source code distributions retain the above copyright notice and this paragraph in its entirety, (2) distributions including binary code include the above copyright notice and this paragraph in its entirety in the documentation or other materials provided with the distribution, and (3) all advertising materials mentioning features or use of this software display the following acknowledgement: ``This product includes software developed by the University of California, Lawrence Berkeley Laboratory and its contributors.'' Neither the name of the University nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission. THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. Enjoy. - Van Jacobson, Steve McCanne, Craig Leres ---------------------------- This directory also contains some short awk programs intended as examples of ways to reduce tcpdump data when you're tracking particular network problems: send-ack.awk Simplifies the tcpdump trace for an ftp (or other unidirectional tcp transfer). Since we assume that one host only sends and the other only acks, all address information is left off and we just note if the packet is a "send" or an "ack". There is one output line per line of the original trace. Field 1 is the packet time in decimal seconds, relative to the start of the conversation. Field 2 is delta-time from last packet. Field 3 is packet type/direction. "Send" means data going from sender to receiver, "ack" means an ack going from the receiver to the sender. A preceding "*" indicates that the data is a retransmission. A preceding "-" indicates a hole in the sequence space (i.e., missing packet(s)), a "#" means an odd-size (not max seg size) packet. Field 4 has the packet flags (same format as raw trace). Field 5 is the sequence number (start seq. num for sender, next expected seq number for acks). The number in parens following an ack is the delta-time from the first send of the packet to the ack. A number in parens following a send is the delta-time from the first send of the packet to the current send (on duplicate packets only). Duplicate sends or acks have a number in square brackets showing the number of duplicates so far. Here is a short sample from near the start of an ftp: 3.00 0.20 send . 512 3.20 0.20 ack . 1024 (0.20) 3.20 0.00 send P 1024 3.40 0.20 ack . 1536 (0.20) 3.80 0.40 * send . 0 (3.80) [2] 3.82 0.02 * ack . 1536 (0.62) [2] Three seconds into the conversation, bytes 512 through 1023 were sent. 200ms later they were acked. Shortly thereafter bytes 1024-1535 were sent and again acked after 200ms. Then, for no apparent reason, 0-511 is retransmitted, 3.8 seconds after its initial send (the round trip time for this ftp was 1sec, +-500ms). Since the receiver is expecting 1536, 1536 is re-acked when 0 arrives. packetdat.awk Computes chunk summary data for an ftp (or similar unidirectional tcp transfer). [A "chunk" refers to a chunk of the sequence space -- essentially the packet sequence number divided by the max segment size.] A summary line is printed showing the number of chunks, the number of packets it took to send that many chunks (if there are no lost or duplicated packets, the number of packets should equal the number of chunks) and the number of acks. Following the summary line is one line of information per chunk. The line contains eight fields: 1 - the chunk number 2 - the start sequence number for this chunk 3 - time of first send 4 - time of last send 5 - time of first ack 6 - time of last ack 7 - number of times chunk was sent 8 - number of times chunk was acked (all times are in decimal seconds, relative to the start of the conversation.) As an example, here is the first part of the output for an ftp trace: # 134 chunks. 536 packets sent. 508 acks. 1 1 0.00 5.80 0.20 0.20 4 1 2 513 0.28 6.20 0.40 0.40 4 1 3 1025 1.16 6.32 1.20 1.20 4 1 4 1561 1.86 15.00 2.00 2.00 6 1 5 2049 2.16 15.44 2.20 2.20 5 1 6 2585 2.64 16.44 2.80 2.80 5 1 7 3073 3.00 16.66 3.20 3.20 4 1 8 3609 3.20 17.24 3.40 5.82 4 11 9 4097 6.02 6.58 6.20 6.80 2 5 This says that 134 chunks were transfered (about 70K since the average packet size was 512 bytes). It took 536 packets to transfer the data (i.e., on the average each chunk was transmitted four times). Looking at, say, chunk 4, we see it represents the 512 bytes of sequence space from 1561 to 2048. It was first sent 1.86 seconds into the conversation. It was last sent 15 seconds into the conversation and was sent a total of 6 times (i.e., it was retransmitted every 2 seconds on the average). It was acked once, 140ms after it first arrived. stime.awk atime.awk Output one line per send or ack, respectively, in the form <time> <seq. number> where <time> is the time in seconds since the start of the transfer and <seq. number> is the sequence number being sent or acked. I typically plot this data looking for suspicious patterns. The problem I was looking at was the bulk-data-transfer throughput of medium delay network paths (1-6 sec. round trip time) under typical DARPA Internet conditions. The trace of the ftp transfer of a large file was used as the raw data source. The method was: - On a local host (but not the Sun running tcpdump), connect to the remote ftp. - On the monitor Sun, start the trace going. E.g., tcpdump host local-host and remote-host and port ftp-data >tracefile - On local, do either a get or put of a large file (~500KB), preferably to the null device (to minimize effects like closing the receive window while waiting for a disk write). - When tranfer is finished, stop tcpdump. Use awk to make up two files of summary data (maxsize is the maximum packet size, tracedata is the file of tcpdump tracedata): awk -f send-ack.awk packetsize=avgsize tracedata >sa awk -f packetdat.awk packetsize=avgsize tracedata >pd - While the summary data files are printing, take a look at how the transfer behaved: awk -f stime.awk tracedata | xgraph (90% of what you learn seems to happen in this step). - Do all of the above steps several times, both directions, at different times of day, with different protocol implementations on the other end. - Using one of the Unix data analysis packages (in my case, S and Gary Perlman's Unix|Stat), spend a few months staring at the data. - Change something in the local protocol implementation and redo the steps above. - Once a week, tell your funding agent that you're discovering wonderful things and you'll write up that research report "real soon now".