Application Layer Protocols
TCP/IP is using some of the Application layer protocols in addition to Transport and Internet layer protocols. Those protocols are used for transforming user requests to network friendly format through the TCP or UDP with IP.
It is protocol used for remote login through the TCP/IP. It is using port 23 for the connection and behavior on the remote machine is the same as on users machine. This protocol is commonly used for interconnecting computers with different systems like PC’s and UNIX. Telnet has benefits, like ability to control host over the network, but it is extremely insecure and should not be used over public or unprotected network. For that scenario new methods were developed. One of the most popular alternatives is SSH.
FTP (File Transfer Protocol) is used for sending and receiving files through the ports 20 and 21 using TCP/IP. For successful transfer, the host should have the FTP server installed, while client – FTP client. FTP client is built in most OS systems, while not always providing visual user interface. Data can be exchanged through the TCP using command prompt. When user tries to access FTP server on the host machine, it will respond with greeting if the server is running.
Some of the FTP servers require logon, but if the server is used for software update it can accept anonymous logon if it is specified in server settings. However, private FTP servers would require legit login and password in order to access its contents.
With the development of FTP clients, it is not necessary to remember FTP commands – most of modern Web-browsers can access FTP server in a useful browser-like manner.
FTP has been developing over the years, and it got more secure version called Secure FTP or SFTP.
TFTP (Trivial File Transfer Protocol) is a simplified version of FTP. It is used for file transfer, but unlike FTP it uses port 69 and UDP instead of TCP. It does not require user log on to the host, but it also does not guarantee data delivery. Transferring data starts after client received response from remote host for the TFTP request. Also, TFTP does not provide host server directory browsing.
Diskless workstations (machines with no hard drive whatsoever) would benefit from TFTP as it allows to load program files from the host directly in workstation’s memory. After work is done and program is closed all the data will be released leaving nothing left in workstation’s memory. However, it is clear that this protocol is very insecure since it requires no user identification, and should be monitored by system administrators very carefully.
Many network protocols rely on time. Timing is critical for routing, time-stamped security methods, accuracy and consistency maintenance. Timing is synchronized in the network using NTP (Network Time Protocol) that communicates with network through port 123. It relies on UDP and its connectionless nature at the Transport layer.
Unlike previous examples PING is not a protocol, but it is an application layer utility. PING (Packet Internet Groper) is useful for determining if the TCP/IP is running correctly, if the NIC is working properly, if the Internet or LAN is accessible through the network. Pinging – means sending echo request to the machine (it can be the same machine where it is sent from) and waits for the echo reply to the sender.
Most of the websites can be accessed either through the web-site name or through its IP. Pinging can be done through the command prompt. Simple ping www.example.com is enough to do the process. It can also be done by using IP instead of website name like ping 192.168.12.24. If the website is up and running, and users computer is connected to the network, user will receive response that include multiple replies. If something is not working (web-site, network configuration, NIC) user would get “request timed out” response. In some cases, everything is working, but reply is still timed out from the web-site. It can happen if organisation managing web-site prevents its equipment from accepting or responding PING commands. Examples of successful and unsuccessful ping could be seen in Figure 1.
Figure 1. Output of successful and unsuccessful PING commands in Windows prompt.
If client’s workstation cannot access Internet, following steps should be implemented:
Check if the local machines TCP/IP is running properly. It can be done using loopback address (127.0.0.1) for the ping. Failing to get response here means that TCP/IP should be reconfigured (or the NIC is broken).
Ping neighbor’s machine in the network. If it fails, then something is wrong with the network connection.
Ping machine in another subnet. If it does not give positive response, then either TCP/IP configuration is having wrong configuration, or there is faulty equipment in the network between user’s machine and another subnet.