Index 1. General Questions: 1.1 What is Wireshark? 1.2 What's up with the name change? Is Wireshark a fork? 1.3 Where can I get help? 1.4 What kind of shark is Wireshark? 1.5 How is Wireshark pronounced, spelled and capitalized? 1.6 How much does Wireshark cost? 1.7 But I just paid someone on eBay for a copy of Wireshark! Did I get ripped off? 1.8 Can I use Wireshark commercially? 1.9 Can I use Wireshark as part of my commercial product? 1.10 What protocols are currently supported? 1.11 Are there any plans to support {your favorite protocol}? 1.12 Can Wireshark read capture files from {your favorite network analyzer}? 1.13 What devices can Wireshark use to capture packets? 1.14 Does Wireshark work on Windows Vista or Windows Server 2008? 2. Downloading Wireshark: 2.1 Why do I get an error when I try to run the Win32 installer? 3. Installing Wireshark: 3.1 I installed the Wireshark RPM (or other package); why did it install TShark but not Wireshark? 4. Building Wireshark: 4.1 I have libpcap installed; why did the configure script not find pcap.h or bpf.h? 4.2 Why do I get the error dftest_DEPENDENCIES was already defined in condition TRUE, which implies condition HAVE_PLUGINS_TRUE when I try to build Wireshark from SVN or a SVN snapshot? 4.3 Why does the linker fail with a number of "Output line too long." messages followed by linker errors when I try to build Wireshark? 4.4 When I try to build Wireshark on Solaris, why does the link fail complaining that plugin_list is undefined? 4.5 When I try to build Wireshark on Windows, why does the build fail because of conflicts between winsock.h and winsock2.h? 5. Starting Wireshark: 5.1 Why does Wireshark crash with a Bus Error when I try to run it on Solaris 8? 5.2 When I run Wireshark on Windows NT, why does it die with a Dr. Watson error, reporting an "Integer division by zero" exception, when I start it? 5.3 When I try to run Wireshark, why does it complain about sprint_realloc_objid being undefined? 5.4 I've installed Wireshark from Fink on Mac OS X; why is it very slow to start up? 6. Crashes and other fatal errors: 6.1 I have an XXX network card on my machine; if I try to capture on it, why does my machine crash or reset itself? 6.2 Why does my machine crash or reset itself when I select "Start" from the "Capture" menu or select "Preferences" from the "Edit" menu? 7. Capturing packets: 7.1 When I use Wireshark to capture packets, why do I see only packets to and from my machine, or not see all the traffic I'm expecting to see from or to the machine I'm trying to monitor? 7.2 When I capture with Wireshark, why can't I see any TCP packets other than packets to and from my machine, even though another analyzer on the network sees those packets? 7.3 Why am I only seeing ARP packets when I try to capture traffic? 7.4 Why am I not seeing any traffic when I try to capture traffic? 7.5 Can Wireshark capture on (my T1/E1 line, SS7 links, etc.)? 7.6 How do I put an interface into promiscuous mode? 7.7 I can set a display filter just fine; why don't capture filters work? 7.8 I'm entering valid capture filters; why do I still get "parse error" errors? 7.9 How can I capture packets with CRC errors? 7.10 How can I capture entire frames, including the FCS? 7.11 I'm capturing packets on a machine on a VLAN; why don't the packets I'm capturing have VLAN tags? 7.12 Why does Wireshark hang after I stop a capture? 8. Capturing packets on Windows: 8.1 I'm running Wireshark on Windows; why does some network interface on my machine not show up in the list of interfaces in the "Interface:" field in the dialog box popped up by "Capture->Start", and/or why does Wireshark give me an error if I try to capture on that interface? 8.2 I'm running Wireshark on Windows; why do no network interfaces show up in the list of interfaces in the "Interface:" field in the dialog box popped up by "Capture->Start"? 8.3 I'm running Wireshark on Windows; why doesn't my serial port/ADSL modem/ISDN modem show up in the list of interfaces in the "Interface:" field in the dialog box popped up by "Capture->Start"? 8.4 I'm running Wireshark on Windows NT 4.0/Windows 2000/Windows XP/Windows Server 2003; my machine has a PPP (dial-up POTS, ISDN, etc.) interface, and it shows up in the "Interface" item in the "Capture Options" dialog box. Why can no packets be sent on or received from that network while I'm trying to capture traffic on that interface? 8.5 I'm running Wireshark on Windows; why am I not seeing any traffic being sent by the machine running Wireshark? 8.6 When I capture on Windows in promiscuous mode, I can see packets other than those sent to or from my machine; however, those packets show up with a "Short Frame" indication, unlike packets to or from my machine. What should I do to arrange that I see those packets in their entirety? 8.7 I'm trying to capture 802.11 traffic on Windows; why am I not seeing any packets? 8.8 I'm trying to capture 802.11 traffic on Windows; why am I seeing packets received by the machine on which I'm capturing traffic, but not packets sent by that machine? 8.9 I'm trying to capture Ethernet VLAN traffic on Windows, and I'm capturing on a "raw" Ethernet device rather than a "VLAN interface", so that I can see the VLAN headers; why am I seeing packets received by the machine on which I'm capturing traffic, but not packets sent by that machine? 9. Capturing packets on UN*Xes: 9.1 I'm running Wireshark on a UNIX-flavored OS; why does some network interface on my machine not show up in the list of interfaces in the "Interface:" field in the dialog box popped up by "Capture->Start", and/or why does Wireshark give me an error if I try to capture on that interface? 9.2 I'm running Wireshark on a UNIX-flavored OS; why do no network interfaces show up in the list of interfaces in the "Interface:" field in the dialog box popped up by "Capture->Start"? 9.3 I'm capturing packets on Linux; why do the time stamps have only 100ms resolution, rather than 1us resolution? 10. Capturing packets on wireless LANs: 10.1 How can I capture raw 802.11 frames, including non-data (management, beacon) frames? 10.2 How do I capture on an 802.11 device in monitor mode? 11. Viewing traffic: 11.1 Why am I seeing lots of packets with incorrect TCP checksums? 11.2 I've just installed Wireshark, and the traffic on my local LAN is boring. Where can I find more interesting captures? 11.3 Why doesn't Wireshark correctly identify RTP packets? It shows them only as UDP. 11.4 Why doesn't Wireshark show Yahoo Messenger packets in captures that contain Yahoo Messenger traffic? 12. Filtering traffic: 12.1 I saved a filter and tried to use its name to filter the display; why do I get an "Unexpected end of filter string" error? 12.2 How can I search for, or filter, packets that have a particular string anywhere in them? 12.3 How do I filter a capture to see traffic for virus XXX? 1. General Questions Q 1.1: What is Wireshark? A: Wireshark(R) is the world's most popular network protocol analyzer. It has a rich and powerful feature set and runs on most computing platforms including Windows, OS X, Linux, and UNIX. Network professionals, security experts, developers, and educators around the world use it regularly. It is freely available as open source, and is released under the GNU General Public License version 2. It is developed and maintained by a global team of protocol experts, and it is an example of a disruptive technology. Wireshark used to be known as Ethereal(R). See the next question for details about the name change. If you're still using Ethereal, it is strongly recommended that you upgrade to Wireshark. For more information, please see the About Wireshark page. Q 1.2: What's up with the name change? Is Wireshark a fork? A: In May of 2006, Gerald Combs (the original author of Ethereal) went to work for CACE Technologies (best known for WinPcap). Unfortunately, he had to leave the Ethereal trademarks behind. This left the project in an awkward position. The only reasonable way to ensure the continued success of the project was to change the name. This is how Wireshark was born. Wireshark is almost (but not quite) a fork. Normally a "fork" of an open source project results in two names, web sites, development teams, support infrastructures, etc. This is the case with Wireshark except for one notable exception -- every member of the core development team is now working on Wireshark. There has been no active development on Ethereal since the name change. Several parts of the Ethereal web site (such as the mailing lists, source code repository, and build farm) have gone offline. More information on the name change can be found here: * * * Many other articles in Q 1.3: Where can I get help? A: Community support is available on the wireshark-users mailing list. Subscription information and archives for all of Wireshark's mailing lists can be found at https://www.wireshark.org/mailman/listinfo. An IRC channel dedicated to Wireshark can be found at irc://irc.freenode.net/wireshark. Self-paced and instructor-led training is available at Wireshark University. A certification program will be announced in Q3 2007. Commercial support and development services are available from CACE Technologies. Q 1.4: What kind of shark is Wireshark? A: carcharodon photoshopia. Q 1.5: How is Wireshark pronounced, spelled and capitalized? A: Wireshark is pronounced as the word wire followed immediately by the word shark. Exact pronunciation and emphasis may vary depending on your locale (e.g. Arkansas). It's spelled with a capital W, followed by a lower-case ireshark. It is not a CamelCase word, i.e., WireShark is incorrect. Q 1.6: How much does Wireshark cost? A: Wireshark is "free software"; you can download it without paying any license fee. The version of Wireshark you download isn't a "demo" version, with limitations not present in a "full" version; it is the full version. The license under which Wireshark is issued is the GNU General Public License version 2. See the GNU GPL FAQ for some more information. Q 1.7: But I just paid someone on eBay for a copy of Wireshark! Did I get ripped off? A: That depends. Did they provide any sort of value-added product or service, such as installation support, installation media, training, trace file analysis, or funky-colored shark-themed socks? Probably not. Wireshark is available for anyone to download, absolutely free, at any time. Paying for a copy implies that you should get something for your money. Q 1.8: Can I use Wireshark commercially? A: Yes, if, for example, you mean "I work for a commercial organization; can I use Wireshark to capture and analyze network traffic in our company's networks or in our customer's networks?" If you mean "Can I use Wireshark as part of my commercial product?", see the next entry in the FAQ. Q 1.9: Can I use Wireshark as part of my commercial product? A: As noted, Wireshark is licensed under the GNU General Public License. The GPL imposes conditions on your use of GPL'ed code in your own products; you cannot, for example, make a "derived work" from Wireshark, by making modifications to it, and then sell the resulting derived work and not allow recipients to give away the resulting work. You must also make the changes you've made to the Wireshark source available to all recipients of your modified version; those changes must also be licensed under the terms of the GPL. See the GPL FAQ for more details; in particular, note the answer to the question about modifying a GPLed program and selling it commercially, and the question about linking GPLed code with other code to make a proprietary program. You can combine a GPLed program such as Wireshark and a commercial program as long as they communicate "at arm's length", as per this item in the GPL FAQ. We recommend keeping Wireshark and your product completely separate, communicating over sockets or pipes. If you're loading any part of Wireshark as a DLL, you're probably doing it wrong. Q 1.10: What protocols are currently supported? A: There are currently hundreds of supported protocols and media. Details can be found in the wireshark(1) man page. Q 1.11: Are there any plans to support {your favorite protocol}? A: Support for particular protocols is added to Wireshark as a result of people contributing that support; no formal plans for adding support for particular protocols in particular future releases exist. Q 1.12: Can Wireshark read capture files from {your favorite network analyzer}? A: Support for particular protocols is added to Wireshark as a result of people contributing that support; no formal plans for adding support for particular protocols in particular future releases exist. If a network analyzer writes out files in a format already supported by Wireshark (e.g., in libpcap format), Wireshark may already be able to read them, unless the analyzer has added its own proprietary extensions to that format. If a network analyzer writes out files in its own format, or has added proprietary extensions to another format, in order to make Wireshark read captures from that network analyzer, we would either have to have a specification for the file format, or the extensions, sufficient to give us enough information to read the parts of the file relevant to Wireshark, or would need at least one capture file in that format AND a detailed textual analysis of the packets in that capture file (showing packet time stamps, packet lengths, and the top-level packet header) in order to reverse-engineer the file format. Note that there is no guarantee that we will be able to reverse-engineer a capture file format. Q 1.13: What devices can Wireshark use to capture packets? A: Wireshark can read live data from Ethernet, Token-Ring, FDDI, serial (PPP and SLIP) (if the OS on which it's running allows Wireshark to do so), 802.11 wireless LAN (if the OS on which it's running allows Wireshark to do so), ATM connections (if the OS on which it's running allows Wireshark to do so), and the "any" device supported on Linux by recent versions of libpcap. See the list of supported capture media on various OSes for details (several items in there say "Unknown", which doesn't mean "Wireshark can't capture on them", it means "we don't know whether it can capture on them"; we expect that it will be able to capture on many of them, but we haven't tried it ourselves - if you try one of those types and it works, please update the wiki page accordingly. It can also read a variety of capture file formats, including: * AG Group/WildPackets EtherPeek/TokenPeek/AiroPeek/EtherHelp/Packet Grabber captures * AIX's iptrace captures * Accellent's 5Views LAN agent output * Cinco Networks NetXRay captures * Cisco Secure Intrusion Detection System IPLog output * CoSine L2 debug output * DBS Etherwatch VMS text output * Endace Measurement Systems' ERF format captures * EyeSDN USB S0 traces * HP-UX nettl captures * ISDN4BSD project i4btrace captures * Linux Bluez Bluetooth stack hcidump -w traces * Lucent/Ascend router debug output * Microsoft Network Monitor captures * Network Associates Windows-based Sniffer captures * Network General/Network Associates DOS-based Sniffer (compressed or uncompressed) captures * Network Instruments Observer version 9 captures * Novell LANalyzer captures * RADCOM's WAN/LAN analyzer captures * Shomiti/Finisar Surveyor captures * Toshiba's ISDN routers dump output * VMS TCPIPtrace/TCPtrace/UCX$TRACE output * Visual Networks' Visual UpTime traffic capture * libpcap, tcpdump and various other tools using tcpdump's capture format * snoop and atmsnoop output so that it can read traces from various network types, as captured by other applications or equipment, even if it cannot itself capture on those network types. Q 1.14: Does Wireshark work on Windows Vista or Windows Server 2008? A: Yes, but if you want to capture packets as a normal user, you must make sure npf.sys is loaded. Wireshark's installer enables this by default. This is not a concern if you run Wireshark as Administrator, but this is discouraged. See the CapturePrivileges page on the wiki for more details. 2. Downloading Wireshark Q 2.1: Why do I get an error when I try to run the Win32 installer? A: The program you used to download it may have downloaded it incorrectly. Web browsers and download accelerators sometimes may do this. Try downloading it with, for example: * Wget, for which Windows binaries are available from Christopher Lewis or wGetGUI, which offers a GUI interface that uses wget; * WS_FTP from Ipswitch, * the ftp command that comes with Windows. If you use the ftp command, make sure you do the transfer in binary mode rather than ASCII mode, by using the binary command before transferring the file. 3. Installing Wireshark Q 3.1: I installed the Wireshark RPM (or other package); why did it install TShark but not Wireshark? A: Many distributions have separate Wireshark packages, one for non-GUI components such as TShark, editcap, dumpcap, etc. and one for the GUI. If this is the case on your system, there's probably a separate package named wireshark-gnome or wireshark-gtk+. Find it and install it. 4. Building Wireshark Q 4.1: I have libpcap installed; why did the configure script not find pcap.h or bpf.h? A: Are you sure pcap.h and bpf.h are installed? The official distribution of libpcap only installs the libpcap.a library file when "make install" is run. To install pcap.h and bpf.h, you must run "make install-incl". If you're running Debian or Redhat, make sure you have the "libpcap-dev" or "libpcap-devel" packages installed. It's also possible that pcap.h and bpf.h have been installed in a strange location. If this is the case, you may have to tweak aclocal.m4. Q 4.2: Why do I get the error dftest_DEPENDENCIES was already defined in condition TRUE, which implies condition HAVE_PLUGINS_TRUE when I try to build Wireshark from SVN or a SVN snapshot? A: You probably have automake 1.5 installed on your machine (the command automake --version will report the version of automake on your machine). There is a bug in that version of automake that causes this problem; upgrade to a later version of automake (1.6 or later). Q 4.3: Why does the linker fail with a number of "Output line too long." messages followed by linker errors when I try to build Wireshark? A: The version of the sed command on your system is incapable of handling very long lines. On Solaris, for example, /usr/bin/sed has a line length limit too low to allow libtool to work; /usr/xpg4/bin/sed can handle it, as can GNU sed if you have it installed. On Solaris, changing your command search path to search /usr/xpg4/bin before /usr/bin should make the problem go away; on any platform on which you have this problem, installing GNU sed and changing your command path to search the directory in which it is installed before searching the directory with the version of sed that came with the OS should make the problem go away. Q 4.4: When I try to build Wireshark on Solaris, why does the link fail complaining that plugin_list is undefined? A: This appears to be due to a problem with some versions of the GTK+ and GLib packages from www.sunfreeware.org; un-install those packages, and try getting the 1.2.10 versions from that site, or the versions from The Written Word, or the versions from Sun's GNOME distribution, or the versions from the supplemental software CD that comes with the Solaris media kit, or build them from source from the GTK Web site. Then re-run the configuration script, and try rebuilding Wireshark. (If you get the 1.2.10 versions from www.sunfreeware.org, and the problem persists, un-install them and try installing one of the other versions mentioned.) Q 4.5: When I try to build Wireshark on Windows, why does the build fail because of conflicts between winsock.h and winsock2.h? A: As of Wireshark 0.9.5, you must install WinPcap 2.3 or later, and the corresponding version of the developer's pack, in order to be able to compile Wireshark; it will not compile with older versions of the developer's pack. The symptoms of this failure are conflicts between definitions in winsock.h and in winsock2.h; Wireshark uses winsock2.h, but pre-2.3 versions of the WinPcap developer's packet use winsock.h. (2.3 uses winsock2.h, so if Wireshark were to use winsock.h, it would not be able to build with current versions of the WinPcap developer's pack.) Note that the installed version of the developer's pack should be the same version as the version of WinPcap you have installed. 5. Starting Wireshark Q 5.1: Why does Wireshark crash with a Bus Error when I try to run it on Solaris 8? A: Some versions of the GTK+ library from www.sunfreeware.org appear to be buggy, causing Wireshark to drop core with a Bus Error. Un-install those packages, and try getting the 1.2.10 version from that site, or the version from The Written Word, or the version from Sun's GNOME distribution, or the version from the supplemental software CD that comes with the Solaris media kit, or build it from source from the GTK Web site. Update the GLib library to the 1.2.10 version, from the same source, as well. (If you get the 1.2.10 versions from www.sunfreeware.org, and the problem persists, un-install them and try installing one of the other versions mentioned.) Similar problems may exist with older versions of GTK+ for earlier versions of Solaris. Q 5.2: When I run Wireshark on Windows NT, why does it die with a Dr. Watson error, reporting an "Integer division by zero" exception, when I start it? A: In at least some case, this appears to be due to using the default VGA driver; if that's not the correct driver for your video card, try running the correct driver for your video card. Q 5.3: When I try to run Wireshark, why does it complain about sprint_realloc_objid being undefined? A: Wireshark can only be linked with version 4.2.2 or later of UCD SNMP. Your version of Wireshark was dynamically linked with such a version of UCD SNMP; however, you have an older version of UCD SNMP installed, which means that when Wireshark is run, it tries to link to the older version, and fails. You will have to replace that version of UCD SNMP with version 4.2.2 or a later version. Q 5.4: I've installed Wireshark from Fink on Mac OS X; why is it very slow to start up? A: When an application is installed on OS X, prior to 10.4, it is usually "prebound" to speed up launching the application. (That's what the "Optimizing" phase of installation is.) Fink normally performs prebinding automatically when you install a package. However, in some rare cases, for whatever reason the prebinding caches get corrupt, and then not only does prebinding fail, but startup actually becomes much slower, because the system tries in vain to perform prebinding "on the fly" as you launch the application. This fails, causing sometimes huge delays. To fix the prebinding caches, run the command sudo /sw/var/lib/fink/prebound/update-package-prebinding.pl -f 6. Crashes and other fatal errors Q 6.1: I have an XXX network card on my machine; if I try to capture on it, why does my machine crash or reset itself? A: This is almost certainly a problem with one or more of: * the operating system you're using; * the device driver for the interface you're using; * the libpcap/WinPcap library and, if this is Windows, the WinPcap device driver; so: * if you are using Windows, see the WinPcap support page - check the "Submitting bugs" section; * if you are using some Linux distribution, some version of BSD, or some other UNIX-flavored OS, you should report the problem to the company or organization that produces the OS (in the case of a Linux distribution, report the problem to whoever produces the distribution). Q 6.2: Why does my machine crash or reset itself when I select "Start" from the "Capture" menu or select "Preferences" from the "Edit" menu? A: Both of those operations cause Wireshark to try to build a list of the interfaces that it can open; it does so by getting a list of interfaces and trying to open them. There is probably an OS, driver, or, for Windows, WinPcap bug that causes the system to crash when this happens; see the previous question. 7. Capturing packets Q 7.1: When I use Wireshark to capture packets, why do I see only packets to and from my machine, or not see all the traffic I'm expecting to see from or to the machine I'm trying to monitor? A: This might be because the interface on which you're capturing is plugged into an Ethernet or Token Ring switch; on a switched network, unicast traffic between two ports will not necessarily appear on other ports - only broadcast and multicast traffic will be sent to all ports. Note that even if your machine is plugged into a hub, the "hub" may be a switched hub, in which case you're still on a switched network. Note also that on the Linksys Web site, they say that their auto-sensing hubs "broadcast the 10Mb packets to the port that operate at 10Mb only and broadcast the 100Mb packets to the ports that operate at 100Mb only", which would indicate that if you sniff on a 10Mb port, you will not see traffic coming sent to a 100Mb port, and vice versa. This problem has also been reported for Netgear dual-speed hubs, and may exist for other "auto-sensing" or "dual-speed" hubs. Some switches have the ability to replicate all traffic on all ports to a single port so that you can plug your analyzer into that single port to sniff all traffic. You would have to check the documentation for the switch to see if this is possible and, if so, to see how to do this. See the switch reference page on the Wireshark Wiki for information on some switches. (Note that it's a Wiki, so you can update or fix that information, or add additional information on those switches or information on new switches, yourself.) Note also that many firewall/NAT boxes have a switch built into them; this includes many of the "cable/DSL router" boxes. If you have a box of that sort, that has a switch with some number of Ethernet ports into which you plug machines on your network, and another Ethernet port used to connect to a cable or DSL modem, you can, at least, sniff traffic between the machines on your network and the Internet by plugging the Ethernet port on the router going to the modem, the Ethernet port on the modem, and the machine on which you're running Wireshark into a hub (make sure it's not a switching hub, and that, if it's a dual-speed hub, all three of those ports are running at the same speed. If your machine is not plugged into a switched network or a dual-speed hub, or it is plugged into a switched network but the port is set up to have all traffic replicated to it, the problem might be that the network interface on which you're capturing doesn't support "promiscuous" mode, or because your OS can't put the interface into promiscuous mode. Normally, network interfaces supply to the host only: * packets sent to one of that host's link-layer addresses; * broadcast packets; * multicast packets sent to a multicast address that the host has configured the interface to accept. Most network interfaces can also be put in "promiscuous" mode, in which they supply to the host all network packets they see. Wireshark will try to put the interface on which it's capturing into promiscuous mode unless the "Capture packets in promiscuous mode" option is turned off in the "Capture Options" dialog box, and TShark will try to put the interface on which it's capturing into promiscuous mode unless the -p option was specified. However, some network interfaces don't support promiscuous mode, and some OSes might not allow interfaces to be put into promiscuous mode. If the interface is not running in promiscuous mode, it won't see any traffic that isn't intended to be seen by your machine. It will see broadcast packets, and multicast packets sent to a multicast MAC address the interface is set up to receive. You should ask the vendor of your network interface whether it supports promiscuous mode. If it does, you should ask whoever supplied the driver for the interface (the vendor, or the supplier of the OS you're running on your machine) whether it supports promiscuous mode with that network interface. In the case of token ring interfaces, the drivers for some of them, on Windows, may require you to enable promiscuous mode in order to capture in promiscuous mode. See the Wireshark Wiki item on Token Ring capturing for details. In the case of wireless LAN interfaces, it appears that, when those interfaces are promiscuously sniffing, they're running in a significantly different mode from the mode that they run in when they're just acting as network interfaces (to the extent that it would be a significant effor for those drivers to support for promiscuously sniffing and acting as regular network interfaces at the same time), so it may be that Windows drivers for those interfaces don't support promiscuous mode. Q 7.2: When I capture with Wireshark, why can't I see any TCP packets other than packets to and from my machine, even though another analyzer on the network sees those packets? A: You're probably not seeing any packets other than unicast packets to or from your machine, and broadcast and multicast packets; a switch will normally send to a port only unicast traffic sent to the MAC address for the interface on that port, and broadcast and multicast traffic - it won't send to that port unicast traffic sent to a MAC address for some other interface - and a network interface not in promiscuous mode will receive only unicast traffic sent to the MAC address for that interface, broadcast traffic, and multicast traffic sent to a multicast MAC address the interface is set up to receive. TCP doesn't use broadcast or multicast, so you will only see your own TCP traffic, but UDP services may use broadcast or multicast so you'll see some UDP traffic - however, this is not a problem with TCP traffic, it's a problem with unicast traffic, as you also won't see all UDP traffic between other machines. I.e., this is probably the same question as this earlier one; see the response to that question. Q 7.3: Why am I only seeing ARP packets when I try to capture traffic? A: You're probably on a switched network, and running Wireshark on a machine that's not sending traffic to the switch and not being sent any traffic from other machines on the switch. ARP packets are often broadcast packets, which are sent to all switch ports. I.e., this is probably the same question as this earlier one; see the response to that question. Q 7.4: Why am I not seeing any traffic when I try to capture traffic? A: Is the machine running Wireshark sending out any traffic on the network interface on which you're capturing, or receiving any traffic on that network, or is there any broadcast traffic on the network or multicast traffic to a multicast group to which the machine running Wireshark belongs? If not, this may just be a problem with promiscuous sniffing, either due to running on a switched network or a dual-speed hub, or due to problems with the interface not supporting promiscuous mode; see the response to this earlier question. Otherwise, on Windows, see the response to this question and, on a UNIX-flavored OS, see the response to this question. Q 7.5: Can Wireshark capture on (my T1/E1 line, SS7 links, etc.)? A: Wireshark can only capture on devices supported by libpcap/WinPcap. On most OSes, only devices that can act as network interfaces of the type that support IP are supported as capture devices for libpcap/WinPcap, although the device doesn't necessarily have to be running as an IP interface in order to support traffic capture. On Linux and FreeBSD, libpcap 0.8 and later support the API for Endace Measurement Systems' DAG cards, so that a system with one of those cards, and its driver and libraries, installed can capture traffic with those cards with libpcap-based applications. You would either have to have a version of Wireshark built with that version of libpcap, or a dynamically-linked version of Wireshark and a shared libpcap library with DAG support, in order to do so with Wireshark. You should ask Endace whether that could be used to capture traffic on, for example, your T1/E1 link. See the SS7 capture setup page on the Wireshark Wiki for current information on capturing SS7 traffic on TDM links. Q 7.6: How do I put an interface into promiscuous mode? A: By not disabling promiscuous mode when running Wireshark or TShark. Note, however, that: * the form of promiscuous mode that libpcap (the library that programs such as tcpdump, Wireshark, etc. use to do packet capture) turns on will not necessarily be shown if you run ifconfig on the interface on a UNIX system; * some network interfaces might not support promiscuous mode, and some drivers might not allow promiscuous mode to be turned on - see this earlier question for more information on that; * the fact that you're not seeing any traffic, or are only seeing broadcast traffic, or aren't seeing any non-broadcast traffic other than traffic to or from the machine running Wireshark, does not mean that promiscuous mode isn't on - see this earlier question for more information on that. I.e., this is probably the same question as this earlier one; see the response to that question. Q 7.7: I can set a display filter just fine; why don't capture filters work? A: Capture filters currently use a different syntax than display filters. Here's the corresponding section from the wireshark(1) man page: "Display filters in Wireshark are very powerful; more fields are filterable in Wireshark than in other protocol analyzers, and the syntax you can use to create your filters is richer. As Wireshark progresses, expect more and more protocol fields to be allowed in display filters. Packet capturing is performed with the pcap library. The capture filter syntax follows the rules of the pcap library. This syntax is different from the display filter syntax." The capture filter syntax used by libpcap can be found in the tcpdump(8) man page. Q 7.8: I'm entering valid capture filters; why do I still get "parse error" errors? A: There is a bug in some versions of libpcap/WinPcap that cause it to report parse errors even for valid expressions if a previous filter expression was invalid and got a parse error. Try exiting and restarting Wireshark; if you are using a version of libpcap/WinPcap with this bug, this will "erase" its memory of the previous parse error. If the capture filter that got the "parse error" now works, the earlier error with that filter was probably due to this bug. The bug was fixed in libpcap 0.6; 0.4[.x] and 0.5[.x] versions of libpcap have this bug, but 0.6[.x] and later versions don't. Versions of WinPcap prior to 2.3 are based on pre-0.6 versions of libpcap, and have this bug; WinPcap 2.3 is based on libpcap 0.6.2, and doesn't have this bug. If you are running Wireshark on a UNIX-flavored platform, run "wireshark -v", or select "About Wireshark..." from the "Help" menu in Wireshark, to see what version of libpcap it's using. If it's not 0.6 or later, you will need either to upgrade your OS to get a later version of libpcap, or will need to build and install a later version of libpcap from the tcpdump.org Web site and then recompile Wireshark from source with that later version of libpcap. If you are running Wireshark on Windows with a pre-2.3 version of WinPcap, you will need to un-install WinPcap and then download and install WinPcap 2.3. Q 7.9: How can I capture packets with CRC errors? A: Wireshark can capture only the packets that the packet capture library - libpcap on UNIX-flavored OSes, and the WinPcap port to Windows of libpcap on Windows - can capture, and libpcap/WinPcap can capture only the packets that the OS's raw packet capture mechanism (or the WinPcap driver, and the underlying OS networking code and network interface drivers, on Windows) will allow it to capture. Unless the OS always supplies packets with errors such as invalid CRCs to the raw packet capture mechanism, or can be configured to do so, invalid CRCs to the raw packet capture mechanism, Wireshark - and other programs that capture raw packets, such as tcpdump - cannot capture those packets. You will have to determine whether your OS needs to be so configured and, if so, can be so configured, configure it if necessary and possible, and make whatever changes to libpcap and the packet capture program you're using are necessary, if any, to support capturing those packets. Most OSes probably do not support capturing packets with invalid CRCs on Ethernet, and probably do not support it on most other link-layer types. Some drivers on some OSes do support it, such as some Ethernet drivers on FreeBSD; in those OSes, you might always get those packets, or you might only get them if you capture in promiscuous mode (you'd have to determine which is the case). Note that libpcap does not currently supply to programs that use it an indication of whether the packet's CRC was invalid (because the drivers themselves do not supply that information to the raw packet capture mechanism); therefore, Wireshark will not indicate which packets had CRC errors unless the FCS was captured (see the next question) and you're using Wireshark 0.9.15 and later, in which case Wireshark will check the CRC and indicate whether it's correct or not. Q 7.10: How can I capture entire frames, including the FCS? A: Wireshark can only capture data that the packet capture library - libpcap on UNIX-flavored OSes, and the WinPcap port to Windows of libpcap on Windows - can capture, and libpcap/WinPcap can capture only the data that the OS's raw packet capture mechanism (or the WinPcap driver, and the underlying OS networking code and network interface drivers, on Windows) will allow it to capture. For any particular link-layer network type, unless the OS supplies the FCS of a frame as part of the frame, or can be configured to do so, Wireshark - and other programs that capture raw packets, such as tcpdump - cannot capture the FCS of a frame. You will have to determine whether your OS needs to be so configured and, if so, can be so configured, configure it if necessary and possible, and make whatever changes to libpcap and the packet capture program you're using are necessary, if any, to support capturing the FCS of a frame. Most OSes do not support capturing the FCS of a frame on Ethernet, and probably do not support it on most other link-layer types. Some drivres on some OSes do support it, such as some (all?) Ethernet drivers on NetBSD and possibly the driver for Apple's gigabit Ethernet interface in Mac OS X; in those OSes, you might always get the FCS, or you might only get the FCS if you capture in promiscuous mode (you'd have to determine which is the case). Versions of Wireshark prior to 0.9.15 will not treat an Ethernet FCS in a captured packet as an FCS. 0.9.15 and later will attempt to determine whether there's an FCS at the end of the frame and, if it thinks there is, will display it as such, and will check whether it's the correct CRC-32 value or not. Q 7.11: I'm capturing packets on a machine on a VLAN; why don't the packets I'm capturing have VLAN tags? A: You might be capturing on what might be called a "VLAN interface" - the way a particular OS makes VLANs plug into the networking stack might, for example, be to have a network device object for the physical interface, which takes VLAN packets, strips off the VLAN header and constructs an Ethernet header, and passes that packet to an internal network device object for the VLAN, which then passes the packets onto various higher-level protocol implementations. In order to see the raw Ethernet packets, rather than "de-VLANized" packets, you would have to capture not on the virtual interface for the VLAN, but on the interface corresponding to the physical network device, if possible. See the Wireshark Wiki item on VLAN capturing for details. Q 7.12: Why does Wireshark hang after I stop a capture? A: The most likely reason for this is that Wireshark is trying to look up an IP address in the capture to convert it to a name (so that, for example, it can display the name in the source address or destination address columns), and that lookup process is taking a very long time. Wireshark calls a routine in the OS of the machine on which it's running to convert of IP addresses to the corresponding names. That routine probably does one or more of: * a search of a system file listing IP addresses and names; * a lookup using DNS; * on UNIX systems, a lookup using NIS; * on Windows systems, a NetBIOS-over-TCP query. If a DNS server that's used in an address lookup is not responding, the lookup will fail, but will only fail after a timeout while the system routine waits for a reply. In addition, on Windows systems, if the DNS lookup of the address fails, either because the server isn't responding or because there are no records in the DNS that could be used to map the address to a name, a NetBIOS-over-TCP query will be made. That query involves sending a message to the NetBIOS-over-TCP name service on that machine, asking for the name and other information about the machine. If the machine isn't running software that responds to those queries - for example, many non-Windows machines wouldn't be running that software - the lookup will only fail after a timeout. Those timeouts can cause the lookup to take a long time. If you disable network address-to-name translation - for example, by turning off the "Enable network name resolution" option in the "Capture Options" dialog box for starting a network capture - the lookups of the address won't be done, which may speed up the process of reading the capture file after the capture is stopped. You can make that setting the default by selecting "Preferences" from the "Edit" menu, turning off the "Enable network name resolution" option in the "Name resolution" options in the preferences disalog box, and using the "Save" button in that dialog box; note that this will save all your current preference settings. If Wireshark hangs when reading a capture even with network name resolution turned off, there might, for example, be a bug in one of Wireshark's dissectors for a protocol causing it to loop infinitely. If you're not running the most recent release of Wireshark, you should first upgrade to that release, as, if there's a bug of that sort, it might've been fixed in a release after the one you're running. If the hang occurs in the most recent release of Wireshark, the bug should be reported to the Wireshark developers' mailing list at wireshark-dev@wireshark.org. On UNIX-flavored OSes, please try to force Wireshark to dump core, by sending it a SIGABRT signal (usually signal 6) with the kill command, and then get a stack trace if you have a debugger installed. A stack trace can be obtained by using your debugger (gdb in this example), the Wireshark binary, and the resulting core file. Here's an example of how to use the gdb command backtrace to do so. $ gdb wireshark core (gdb) backtrace ..... prints the stack trace (gdb) quit $ The core dump file may be named "wireshark.core" rather than "core" on some platforms (e.g., BSD systems). Also, if at all possible, please send a copy of the capture file that caused the problem; when capturing packets, Wireshark normally writes captured packets to a temporary file, which will probably be in /tmp or /var/tmp on UNIX-flavored OSes, \TEMP on the main system disk (normally C:) on Windows 9x/Me/NT 4.0, and \Documents and Settings\your login name\Local Settings\Temp on the main system disk on Windows 2000/Windows XP/Windows Server 2003, so the capture file will probably be there. It will have a name beginning with ether, with some mixture of letters and numbers after that. Please don't send a trace file greater than 1 MB when compressed; instead, make it available via FTP or HTTP, or say it's available but leave it up to a developer to ask for it. If the trace file contains sensitive information (e.g., passwords), then please do not send it. 8. Capturing packets on Windows Q 8.1: I'm running Wireshark on Windows; why does some network interface on my machine not show up in the list of interfaces in the "Interface:" field in the dialog box popped up by "Capture->Start", and/or why does Wireshark give me an error if I try to capture on that interface? A: If you are running Wireshark on Windows NT 4.0, Windows 2000, Windows XP, or Windows Server 2003, and this is the first time you have run a WinPcap-based program (such as Wireshark, or TShark, or WinDump, or Analyzer, or...) since the machine was rebooted, you need to run that program from an account with administrator privileges; once you have run such a program, you will not need administrator privileges to run any such programs until you reboot. If you are running on Windows Windows 2000/Windows XP/Windows Server 2003 and have administrator privileges or a WinPcap-based program has been run with those privileges since the machine rebooted, this problem might clear up if you completely un-install WinPcap and then re-install it. If that doesn't work, then note that Wireshark relies on the WinPcap library, on the WinPcap device driver, and on the facilities that come with the OS on which it's running in order to do captures. Therefore, if the OS, the WinPcap library, or the WinPcap driver don't support capturing on a particular network interface device, Wireshark won't be able to capture on that device. Note that: 1. 2.02 and earlier versions of the WinPcap driver and library that Wireshark uses for packet capture didn't support Token Ring interfaces; versions 2.1 and later support Token Ring, and the current version of Wireshark works with (and, in fact, requires) WinPcap 2.1 or later. If you are having problems capturing on Token Ring interfaces, and you have WinPcap 2.02 or an earlier version of WinPcap installed, you should uninstall WinPcap, download and install the current version of WinPcap, and then install the latest version of Wireshark. 2. WinPcap 2.3 has problems supporting PPP WAN interfaces on Windows NT 4.0, Windows 2000, Windows XP, and Windows Server 2003, and, to avoid those problems, support for PPP WAN interfaces on those versions of Windows has been disabled in WinPcap 3.0. Regular dial-up lines, ISDN lines, ADSL connections using PPPoE or PPPoA, and various other lines such as T1/E1 lines are all PPP interfaces, so those interfaces might not show up on the list of interfaces in the "Capture Options" dialog on those OSes. On Windows 2000, Windows XP, and Windows Server 2003, but not Windows NT 4.0 or Windows Vista Beta 1, you should be able to capture on the "GenericDialupAdapter" with WinPcap 3.1. (3.1 beta releases called it the "NdisWanAdapter"; if you're using a 3.1 beta release, you should un-install it and install the final 3.1 release.) See the Wireshark Wiki item on PPP capturing for details. 3. WinPcap prior to 3.0 does not support multiprocessor machines (note that machines with a single multi-threaded processor, such as Intel's new multi-threaded x86 processors, are multiprocessor machines as far as the OS and WinPcap are concerned), and recent 2.x versions of WinPcap refuse to operate if they detect that they're running on a multiprocessor machine, which means that they may not show any network interfaces. You will need to use WinPcap 3.0 to capture on a multiprocessor machine. If an interface doesn't show up in the list of interfaces in the "Interface:" field, and you know the name of the interface, try entering that name in the "Interface:" field and capturing on that device. If the attempt to capture on it succeeds, the interface is somehow not being reported by the mechanism Wireshark uses to get a list of interfaces. Try listing the interfaces with WinDump; see the WinDump Web site for information on using WinDump. You would run WinDump with the -D flag; if it lists the interface, please report this to wireshark-dev@wireshark.org giving full details of the problem, including * the operating system you're using, and the version of that operating system; * the type of network device you're using; * the output of WinDump. If WinDump does not list the interface, this is almost certainly a problem with one or more of: * the operating system you're using; * the device driver for the interface you're using; * the WinPcap library and/or the WinPcap device driver; so first check the WinPcap FAQ or the Wiretapped.net mirror of that FAQ, to see if your problem is mentioned there. If not, then see the WinPcap support page - check the "Submitting bugs" section. If you are having trouble capturing on a particular network interface, first try capturing on that device with WinDump; see the WinDump Web site for information on using WinDump. If you can capture on the interface with WinDump, send mail to wireshark-users@wireshark.org giving full details of the problem, including * the operating system you're using, and the version of that operating system; * the type of network device you're using; * the error message you get from Wireshark. If you cannot capture on the interface with WinDump, this is almost certainly a problem with one or more of: * the operating system you're using; * the device driver for the interface you're using; * the WinPcap library and/or the WinPcap device driver; so first check the WinPcap FAQ or the Wiretapped.net mirror of that FAQ, to see if your problem is mentioned there. If not, then see the WinPcap support page - check the "Submitting bugs" section. You may also want to ask the wireshark-users@wireshark.org and the winpcap-users@winpcap.org mailing lists to see if anybody happens to know about the problem and know a workaround or fix for the problem. (Note that you will have to subscribe to that list in order to be allowed to mail to it; see the WinPcap support page for information on the mailing list.) In your mail, please give full details of the problem, as described above, and also indicate that the problem occurs with WinDump, not just with Wireshark. Q 8.2: I'm running Wireshark on Windows; why do no network interfaces show up in the list of interfaces in the "Interface:" field in the dialog box popped up by "Capture->Start"? A: This is really the same question as a previous one; see the response to that question. Q 8.3: I'm running Wireshark on Windows; why doesn't my serial port/ADSL modem/ISDN modem show up in the list of interfaces in the "Interface:" field in the dialog box popped up by "Capture->Start"? A: Internet access on those devices is often done with the Point-to-Point (PPP) protocol; WinPcap 2.3 has problems supporting PPP WAN interfaces on Windows NT 4.0, Windows 2000, Windows XP, and Windows Server 2003, and, to avoid those problems, support for PPP WAN interfaces on those versions of Windows has been disabled in WinPcap 3.0. On Windows 2000, Windows XP, and Windows Server 2003, but not Windows NT 4.0 or Windows Vista Beta 1, you should be able to capture on the "GenericDialupAdapter" with WinPcap 3.1. (3.1 beta releases called it the "NdisWanAdapter"; if you're using a 3.1 beta release, you should un-install it and install the final 3.1 release.) See the Wireshark Wiki item on PPP capturing for details. Q 8.4: I'm running Wireshark on Windows NT 4.0/Windows 2000/Windows XP/Windows Server 2003; my machine has a PPP (dial-up POTS, ISDN, etc.) interface, and it shows up in the "Interface" item in the "Capture Options" dialog box. Why can no packets be sent on or received from that network while I'm trying to capture traffic on that interface? A: Some versions of WinPcap have problems with PPP WAN interfaces on Windows NT 4.0, Windows 2000, Windows XP, and Windows Server 2003; one symptom that may be seen is that attempts to capture in promiscuous mode on the interface cause the interface to be incapable of sending or receiving packets. You can disable promiscuous mode using the -p command-line flag or the item in the "Capture Preferences" dialog box, but this may mean that outgoing packets, or incoming packets, won't be seen in the capture. On Windows 2000, Windows XP, and Windows Server 2003, but not Windows NT 4.0 or Windows Vista Beta 1, you should be able to capture on the "GenericDialupAdapter" with WinPcap 3.1. (3.1 beta releases called it the "NdisWanAdapter"; if you're using a 3.1 beta release, you should un-install it and install the final 3.1 release.) See the Wireshark Wiki item on PPP capturing for details. Q 8.5: I'm running Wireshark on Windows; why am I not seeing any traffic being sent by the machine running Wireshark? A: If you are running some form of VPN client software, it might be causing this problem; people have seen this problem when they have Check Point's VPN software installed on their machine. If that's the cause of the problem, you will have to remove the VPN software in order to have Wireshark (or any other application using WinPcap) see outgoing packets; unfortunately, neither we nor the WinPcap developers know any way to make WinPcap and the VPN software work well together. Also, some drivers for Windows (especially some wireless network interface drivers) apparently do not, when running in promiscuous mode, arrange that outgoing packets are delivered to the software that requested that the interface run promiscuously; try turning promiscuous mode off. Q 8.6: When I capture on Windows in promiscuous mode, I can see packets other than those sent to or from my machine; however, those packets show up with a "Short Frame" indication, unlike packets to or from my machine. What should I do to arrange that I see those packets in their entirety? A: In at least some cases, this appears to be the result of PGPnet running on the network interface on which you're capturing; turn it off on that interface. Q 8.7: I'm trying to capture 802.11 traffic on Windows; why am I not seeing any packets? A: At least some 802.11 card drivers on Windows appear not to see any packets if they're running in promiscuous mode. Try turning promiscuous mode off; you'll only be able to see packets sent by and received by your machine, not third-party traffic, and it'll look like Ethernet traffic and won't include any management or control frames, but that's a limitation of the card drivers. See MicroLogix's list of cards supported with WinPcap for information on support of various adapters and drivers with WinPcap. Q 8.8: I'm trying to capture 802.11 traffic on Windows; why am I seeing packets received by the machine on which I'm capturing traffic, but not packets sent by that machine? A: This appears to be another problem with promiscuous mode; try turning it off. Q 8.9: I'm trying to capture Ethernet VLAN traffic on Windows, and I'm capturing on a "raw" Ethernet device rather than a "VLAN interface", so that I can see the VLAN headers; why am I seeing packets received by the machine on which I'm capturing traffic, but not packets sent by that machine? A: The way the Windows networking code works probably means that packets are sent on a "VLAN interface" rather than the "raw" device, so packets sent by the machine will only be seen when you capture on the "VLAN interface". If so, you will be unable to see outgoing packets when capturing on the "raw" device, so you are stuck with a choice between seeing VLAN headers and seeing outgoing packets. 9. Capturing packets on UN*Xes Q 9.1: I'm running Wireshark on a UNIX-flavored OS; why does some network interface on my machine not show up in the list of interfaces in the "Interface:" field in the dialog box popped up by "Capture->Start", and/or why does Wireshark give me an error if I try to capture on that interface? A: You may need to run Wireshark from an account with sufficient privileges to capture packets, such as the super-user account, or may need to give your account sufficient privileges to capture packets. Only those interfaces that Wireshark can open for capturing show up in that list; if you don't have sufficient privileges to capture on any interfaces, no interfaces will show up in the list. See the Wireshark Wiki item on capture privileges for details on how to give a particular account or account group capture privileges on platforms where that can be done. If you are running Wireshark from an account with sufficient privileges, then note that Wireshark relies on the libpcap library, and on the facilities that come with the OS on which it's running in order to do captures. On some OSes, those facilities aren't present by default; see the Wireshark Wiki item on adding capture support for details. And, even if you're running with an account that has sufficient privileges to capture, and capture support is present in your OS, if the OS or the libpcap library don't support capturing on a particular network interface device or particular types of devices, Wireshark won't be able to capture on that device. On Solaris, note that libpcap 0.6.2 and earlier didn't support Token Ring interfaces; the current version, 0.7.2, does support Token Ring, and the current version of Wireshark works with libpcap 0.7.2 and later. If an interface doesn't show up in the list of interfaces in the "Interface:" field, and you know the name of the interface, try entering that name in the "Interface:" field and capturing on that device. If the attempt to capture on it succeeds, the interface is somehow not being reported by the mechanism Wireshark uses to get a list of interfaces; please report this to wireshark-dev@wireshark.org giving full details of the problem, including * the operating system you're using, and the version of that operating system (for Linux, give both the version number of the kernel and the name and version number of the distribution you're using); * the type of network device you're using. If you are having trouble capturing on a particular network interface, and you've made sure that (on platforms that require it) you've arranged that packet capture support is present, as per the above, first try capturing on that device with tcpdump. If you can capture on the interface with tcpdump, send mail to wireshark-users@wireshark.org giving full details of the problem, including * the operating system you're using, and the version of that operating system (for Linux, give both the version number of the kernel and the name and version number of the distribution you're using); * the type of network device you're using; * the error message you get from Wireshark. If you cannot capture on the interface with tcpdump, this is almost certainly a problem with one or more of: * the operating system you're using; * the device driver for the interface you're using; * the libpcap library; so you should report the problem to the company or organization that produces the OS (in the case of a Linux distribution, report the problem to whoever produces the distribution). You may also want to ask the wireshark-users@wireshark.org and the tcpdump-workers@lists.tcpdump.org mailing lists to see if anybody happens to know about the problem and know a workaround or fix for the problem. In your mail, please give full details of the problem, as described above, and also indicate that the problem occurs with tcpdump not just with Wireshark. Q 9.2: I'm running Wireshark on a UNIX-flavored OS; why do no network interfaces show up in the list of interfaces in the "Interface:" field in the dialog box popped up by "Capture->Start"? A: This is really the same question as the previous one; see the response to that question. Q 9.3: I'm capturing packets on Linux; why do the time stamps have only 100ms resolution, rather than 1us resolution? A: Wireshark gets time stamps from libpcap/WinPcap, and libpcap/WinPcap get them from the OS kernel, so Wireshark - and any other program using libpcap, such as tcpdump - is at the mercy of the time stamping code in the OS for time stamps. At least on x86-based machines, Linux can get high-resolution time stamps on newer processors with the Time Stamp Counter (TSC) register; for example, Intel x86 processors, starting with the Pentium Pro, and including all x86 processors since then, have had a TSC, and other vendors probably added the TSC at some point to their families of x86 processors. The Linux kernel must be configured with the CONFIG_X86_TSC option enabled in order to use the TSC. Make sure this option is enabled in your kernel. In addition, some Linux distributions may have bugs in their versions of the kernel that cause packets not to be given high-resolution time stamps even if the TSC is enabled. See, for example, bug 61111 for Red Hat Linux 7.2. If your distribution has a bug such as this, you may have to run a standard kernel from kernel.org in order to get high-resolution time stamps. 10. Capturing packets on wireless LANs Q 10.1: How can I capture raw 802.11 frames, including non-data (management, beacon) frames? A: That depends on the operating system on which you're running, and on the 802.11 interface on which you're capturing. This would probably require that you capture in promiscuous mode or in the mode called "monitor mode" or "RFMON mode". On some platforms, or with some cards, this might require that you capture in monitor mode - promiscuous mode might not be sufficient. If you want to capture traffic on networks other than the one with which you're associated, you will have to capture in monitor mode. Not all operating systems support capturing non-data packets and, even on operating systems that do support it, not all drivers, and thus not all interfaces, support it. Even on those that do, monitor mode might not be supported by the operating system or by the drivers for all interfaces. NOTE: an interface running in monitor mode will, on most if not all platforms, not be able to act as a regular network interface; putting it into monitor mode will, in effect, take your machine off of whatever network it's on as long as the interface is in monitor mode, allowing it only to passively capture packets. This means that you should disable name resolution when capturing in monitor mode; otherwise, when Wireshark (or TShark, or tcpdump) tries to display IP addresses as host names, it will probably block for a long time trying to resolve the name because it will not be able to communicate with any DNS or NIS servers. See the Wireshark Wiki item on 802.11 capturing for details. Q 10.2: How do I capture on an 802.11 device in monitor mode? A: Whether you will be able to capture in monitor mode depends on the operating system, adapter, and driver you're using. See the previous question for information on monitor mode, including a link to the Wireshark Wiki page that gives details on 802.11 capturing. 11. Viewing traffic Q 11.1: Why am I seeing lots of packets with incorrect TCP checksums? A: If the packets that have incorrect TCP checksums are all being sent by the machine on which Wireshark is running, this is probably because the network interface on which you're capturing does TCP checksum offloading. That means that the TCP checksum is added to the packet by the network interface, not by the OS's TCP/IP stack; when capturing on an interface, packets being sent by the host on which you're capturing are directly handed to the capture interface by the OS, which means that they are handed to the capture interface without a TCP checksum being added to them. The only way to prevent this from happening would be to disable TCP checksum offloading, but 1. that might not even be possible on some OSes; 2. that could reduce networking performance significantly. However, you can disable the check that Wireshark does of the TCP checksum, so that it won't report any packets as having TCP checksum errors, and so that it won't refuse to do TCP reassembly due to a packet having an incorrect TCP checksum. That can be set as an Wireshark preference by selecting "Preferences" from the "Edit" menu, opening up the "Protocols" list in the left-hand pane of the "Preferences" dialog box, selecting "TCP", from that list, turning off the "Check the validity of the TCP checksum when possible" option, clicking "Save" if you want to save that setting in your preference file, and clicking "OK". It can also be set on the Wireshark or TShark command line with a -o tcp.check_checksum:false command-line flag, or manually set in your preferences file by adding a tcp.check_checksum:false line. Q 11.2: I've just installed Wireshark, and the traffic on my local LAN is boring. Where can I find more interesting captures? A: We have a collection of strange and exotic sample capture files at http://wiki.wireshark.org/SampleCaptures Q 11.3: Why doesn't Wireshark correctly identify RTP packets? It shows them only as UDP. A: Wireshark can identify a UDP datagram as containing a packet of a particular protocol running atop UDP only if 1. The protocol in question has a particular standard port number, and the UDP source or destination port number is that port 2. Packets of that protocol can be identified by looking for a "signature" of some type in the packet - i.e., some data that, if Wireshark finds it in some particular part of a packet, means that the packet is almost certainly a packet of that type. 3. Some other traffic earlier in the capture indicated that, for example, UDP traffic between two particular addresses and ports will be RTP traffic. RTP doesn't have a standard port number, so 1) doesn't work; it doesn't, as far as I know, have any "signature", so 2) doesn't work. That leaves 3). If there's RTSP traffic that sets up an RTP session, then, at least in some cases, the RTSP dissector will set things up so that subsequent RTP traffic will be identified. Currently, that's the only place we do that; there may be other places. However, there will always be places where Wireshark is simply incapable of deducing that a given UDP flow is RTP; a mechanism would be needed to allow the user to specify that a given conversation should be treated as RTP. As of Wireshark 0.8.16, such a mechanism exists; if you select a UDP or TCP packet, the right mouse button menu will have a "Decode As..." menu item, which will pop up a dialog box letting you specify that the source port, the destination port, or both the source and destination ports of the packet should be dissected as some particular protocol. Q 11.4: Why doesn't Wireshark show Yahoo Messenger packets in captures that contain Yahoo Messenger traffic? A: Wireshark only recognizes as Yahoo Messenger traffic packets to or from TCP port 3050 that begin with "YPNS", "YHOO", or "YMSG". TCP segments that start with the middle of a Yahoo Messenger packet that takes more than one TCP segment will not be recognized as Yahoo Messenger packets (even if the TCP segment also contains the beginning of another Yahoo Messenger packet). 12. Filtering traffic Q 12.1: I saved a filter and tried to use its name to filter the display; why do I get an "Unexpected end of filter string" error? A: You cannot use the name of a saved display filter as a filter. To filter the display, you can enter a display filter expression - not the name of a saved display filter - in the "Filter:" box at the bottom of the display, and type the key or press the "Apply" button (that does not require you to have a saved filter), or, if you want to use a saved filter, you can press the "Filter:" button, select the filter in the dialog box that pops up, and press the "OK" button. Q 12.2: How can I search for, or filter, packets that have a particular string anywhere in them? A: If you want to do this when capturing, you can't. That's a feature that would be hard to implement in capture filters without changes to the capture filter code, which, on many platforms, is in the OS kernel and, on other platforms, is in the libpcap library. After capture, you can search for text by selecting Edit->Find Packet... and making sure String is selected. Alternately, you can use the "contains" display filter operator or "matches" operator if it's supported on your system. Q 12.3: How do I filter a capture to see traffic for virus XXX? A: For some viruses/worms there might be a capture filter to recognize the virus traffic. Check the CaptureFilters page on the Wireshark Wiki to see if anybody's added such a filter. Note that Wireshark was not designed to be an intrusion detection system; you might be able to use it as an IDS, but in most cases software designed to be an IDS, such as Snort or Prelude, will probably work better. The Bleeding Edge of Snort has a collection of signatures for Snort to detect various viruses, worms, and the like.