That note mentions both -w and -s 0, as they are very important for getting traces to be sent to somebody else to analyze.This is probably less a wireshark question and more a 'how do I pipe a file into an application' on windows. They should also use -s 0 in the tcpdump command, so that they get the full packet data.Īpple have a pretty good technical note on how to take network traces it discusses this from the point of view of an OS X user, and mentions some OS X-only tools, but it also mentions tcpdump in the 'Getting Started With tcpdump' section, and that section applies to other UN*Xes, once you replace 'If you're running on a system prior to OS X 10.6' with 'If you're using tcpdump 0.x or 1.0.x' and 'on OS X 10.6 and later' with 'with tcpdump 1.1.0 and later', and replace the stuff talking about the -i option with whatever is appropriate for your OS and machine.

At best, you can try to get another trace, if whatever problem you're trying to diagnose can be made to happen again, and this time have them use tcpdump with the -w option, so that it writes out a pcap file. If you need that information in order to solve a problem, you're out of luck.

As the output of tcpdump was its text-mode output, the only information available in the file is the information tcpdump printed even if it were possible to convert that file to a pcap file, the pcap file would not contain any more information than is available in the printout - the TCP payload of the two packets you showed, for example, is permanently lost and you will not ever be able to get it back.

Wireshark tcpdump