TCP/IP troubleshooting

This chapter focuses on problems related to the TCP/IP protocol suite and how to troubleshoot them. TCP/IP protocols are implemented in software and cover the second (Internet), the third (host-to-host), and the upper (application) layers of the ARPA model. These protocols need a network infrastructure as well as a medium in order to communicate. One of the common problems associated with TCP/IP protocol is that the stack not properly installed on local host. The easiest way to confirm this, apart from checking the network configuration via the control panel and visually confirming that TCP/IP is installed for the particular NIC used on the host, is to perform a loop-back test by pinging the host itself. This is done by executing ping localhost or ping 127. 0.0.1. If a response is received, it means that the stack is correctly installed. If it needs to be confirmed that a remote host is available, the particular machine can be checked by pinging it. An incorrect or absent default gateway in the TCP configuration screen means that a host is not able to send a message to another host on a different network. In this case, user can often figure out the problem by simply observing the error messages returned by the ping utility. Before any two devices can communicate using TCP/IP, they need to establish a so-called triple handshake, without which the devices cannot communicate at all. To confirm this, one has to simply try to establish a connection (for example, by using an FTP client to log into an FTP server) and use a protocol analyzer to capture the handshake.