User Datagram Protocol (UDP)

There are several well-known ports for #Transmission Control Protocol (TCP) and #User Datagram Protocol (UDP) authorised by Internet Assigned Numbers Authority (IANA) for #Application Layer protocols ranged from 0 to 1023. They are shown in the following table:

TCP is designed to be connection-oriented and reliable, in contrast to User Datagram Protocol (UDP). This means that connection must be established first (using three-way handshake) via a full-duplex channel and maintained by the protocol during a transmission. The requester will first send out a control bit SYN (stand for synchronisation) and its Initial Sequence Number (ISN) to the server. The server sends back a packet to the requester that contains the control bit ACK (stand for acknowledgement) and its ISN. From now on, the requester will assume the connection is well established, and it will send packets to the server with incremented SYN and ACK (same to the server). This way, TCP can guarantee the eventual orderly delivery of the packets.

Socket programming in User Datagram Protocol (UDP)# is much fewer steps to set up compare to TCP. Both server and client just need to establish socket (pass in SOCK_DGRAM instead of SOCK_STREAM for UDP connection) and bind it to an address for themselves. Then, they could call either sendto() or recvfrom() to send or receive any data transmitted by them. Note that both sendto() and recvfrom() requires to know the destination address. Otherwise, its structure is almost identical to the above TCP implementation.

SNMP supports both central and distributed management (multiple Management Station or Manager with a pool of Management Agent) as each device in the network maintains local Management Information Base (MIB). The communication is usually done in User Datagram Protocol (UDP)# but SNMPv2 allows the use of Transmission Control Protocol (TCP)# too.

#SNMPv1 is defined by RFC 1065 - 1067 which specify as a protocol supporting stand-alone Management Station or Manager and Management Agent. Limited but easily implemented Management Information Base (MIB) of scalar variables and two dimensional tables are defined. It is designed to be a streamlined (simple but efficient) protocol with limited functionality such as allows only centralised network management. The connection between the manager and agent is done using User Datagram Protocol (UDP)#.

RIP will periodically broadcast the router’s routing table to its neighbours’ #port 520 every 30 seconds (default). The RIP message is transferred using #User Datagram Protocol (UDP). There are two types of message: request message and response message. Request message is sent out on startup by RIP enable interface which requests all RIP enabled neighbours to send routing table to it. Response message on the other hand is the response to the request message, which sends the data containing routing table to the requesting interface.

RADIUS is a #User Datagram Protocol (UDP) developed by Livingston Enterprises in mid-1990s which provides centralised #Triple A (AAA) services to their network access server (NAS) devices. It supports Link Encryption# and maximum attribute data size of 255 octets (8 bit). It has implicit support of agent support, meaning the agent behaviours might be implemented in a RADIUS server. RADIUS protocol is standardised by RFC 2138, RFC 2139, RFC 2865, RFC 2866 and RFC 2869.