Configuring Transmission Control Protocol/Internet Protocol (TCP/IP) Even though TCP/IP has become a popular protocol because of its use on the Internet, many people do not understand it. There are some very basic concepts to deal with TCP/IP, and as usual with computers, also some complex areas. Usually, dealing with one item at a time makes TCP/IP simple to understand. First, let’s look at a configuration screen for TCP/IP (this will vary between linux distros): The first item in the configuration is the Connection Name. This is basically the connection on a specific device which you are configuring. In this case, it is called ‘insight_wifi_4163’. Currently, the information for the settings of IPv4 is being modified. The ‘Methods’ are the ways in which the connection is being configured. For the time being let’s focus on the ‘Manual’ setting. NOTE: The other settings allow for automatic settings or even to disable the connection. Three addresses are configured for TCP/IP: Address, Netmask and Gateway. Each device, or host, on a network or even the Internet needs an IP Address. Each IP Address is made up of four bytes. The bytes are usually represented as decimals and separated by dots or periods. The numbers range from 0 to 255 to make up a total of 256 possibilities per byte. As shown in the figure, the IP address for the connection is 10.0.0.154. The IP address is the address seen by the other systems on the Internet so when a response is made to a request, the data can cross from router to router and reach your system. In some cases, the data is sent from systems outside your country, yet they still manage to be delivered. How can this all be managed? Read on for the answers. NOTE: Remember that a host is any device on a network or the Internet which has an IP Address. The next portion of the address is the ‘Netmask’, which is sometimes called the Subnet Mask. The Netmask can be very complicated and is its own article found here. Basically, the Netmask determines which portion of the IP Address is the network ID and the host ID. The next portion of the address is the Gateway. When a host is sending data to another host having an address that does not have the same network ID, the data is forwarded to the Gateway. The Gateway can be a routing or bridging device. For instance, most homes may be connected to the Internet by a wireless hub. The wireless hub is considered your Gateway to Internet. If other systems exist on a LAN, the IP Addresses would have the same network ID, so the Gateway would ignore the network ID of the local network. The next section of the TCP/IP configuration is the Domain Name System (DNS) servers. The addresses given are the IP Addresses of the DNS servers on the Internet. DNS servers manage a database of domain names and their IP Addresses. Domain names are the friendly names of servers on the Internet, such as Linux.org. Linux.org is the friendly name that you type into your browsers address bar. Of course, we know that everything on the Internet with TCP/IP is done with IP Addresses. If your system does not know the IP Address for Linux.org, it sends a request to a DNS server for an IP Address resolution. The first DNS server will be queried, if no response is received in a set amount of time, the secondary DNS server is queried. So, the host queries the DNS server for Linux.org and receives a reply with the address of 126.96.36.199. When packets are sent to Linux.org, the IP Address included in the encapsulated data is the one received from the DNS server. Of course, friendly names cannot be used for the DNS servers since an address is required to make Name Resolution queries. In most cases, systems on a LAN do not have TCP/IP manually configured. Most routing devices, such as wireless hubs can auto-configure a TCP/IP client. The ability to have a device auto-configure a TCP/IP client requires Dynamic Host Configuration Protocol (DHCP) device or server. The device or server is given a range of IP Addresses, a Netmask, Gateway, DNS Server and any other information needed by the network. The device or server tracks the IP Addresses assigned so no duplicates are given. Lease times can also be set so a system must re-lease the address to keep it. By using leases, it is possible to reclaim unused addresses for systems that are no longer in use. NOTE: To enable DHCP configuration, the Method can be changed to DHCP. Now that you understand what the items represent, you may ask, ‘Where do I get an IP Address when I manually set it?’. Most devices which connect a user to the Internet are automatically configured to use a Private Address for the LAN. Private Addresses are IP Addresses which cannot be used on the physical Internet, but only on LANs. The Private Addresses are translated to Public Addresses when the data is transmitted on the Internet. For example, on a network using a private address of 10.0.0.1, the Gateway is 10.0.0.100. The Gateway, say a wireless hub, has two addresses. One address is for the LAN, in this case 10.0.0.100. Another address is used on the side connected to the Internet; in this case, it is 188.8.131.52. When I send a packet to the Internet it goes to my Gateway (10.0.0.100). The device removes the lower layers of encapsulation and replaces the IP Address with its own (184.108.40.206). A special port number is used to signify my system as being the original sender. When data is sent back to the Gateway device (220.127.116.11), it contains the port number sent originally. The port number is checked and found linked to my system. The encapsulated data is removed again and replaced with my IP Address (10.0.0.1). The data is retransmitted on the LAN where my system receives it and handles it accordingly. If you must assign an IP address, use one that is consistent with the Gateway. For more details read the article ‘Netmasking Transmission Control Protocol/Internet Protocol (TCP/IP)’.