Understanding the Session Layer: Communication Protocols
Session Layer: Communication Protocols and Synchronization
The first problem is that your messages can be crossed during the conversation. Both of you can type messages at exactly the same time, interrupting each other. The second problem is the need to stop (to save the current conversation as a file) or verify the previous conversation of each (to get clues of the reason for discussion) or to resynchronize the communication after a break.
- To solve the first problem, you should establish a protocol or a set of protocols, which dictate the rules of communication between you. This means that each of you must agree to a set of rules to be used during the conversation (e.g., the relay when sending messages to prevent an interrupt each other). This is known as alternating two-way communication.
- Another solution is for each individual to type your messages whenever you want, regardless of who is transmitting, and you assume that more information is always coming. This is known as simultaneous two-way communication.
To solve the second problem, one must send the other one a checkpoint, which means that each person should save the conversation. Then, each person should reread the last part of your conversation and check the time on the clock. This is known as synchronization.
Two checkpoints are very important: how the conversation begins and how it ends. This is known as an orderly start and ending the conversation. For example, when you use Instant Mail or Internet Relay Chat, you often dismiss the other person before ending the session. The other person realizes that you’re ending the session.
Analogies of the Session Layer
- The session layer establishes, manages, and terminates sessions between applications. This includes starting, quitting, and re-syncing two computers that are having a “communication session”. The session layer coordinates the applications as they interact on two communicating hosts. Data communication travels over packet-switched networks, unlike phone calls, which travel on circuit-switched networks. Communication between two computers involves many mini-conversations, thus ensuring that two computers can communicate effectively.
- Mini-conversations require that each host perform two functions: to request service, as a client, and respond to the service, as a server. The determination of the role they will perform at a specific time is called dialog control.
Dialog Control
- The session layer decides between using two-way communication simultaneously or alternating two-way communication. This decision is known as dialog control. If simultaneous two-way communication is allowed, the session layer will do little to manage the conversation. In such cases, other layers of the computers that are communicating manage the conversation. You can have bumps in the session layer, although this is very different from the means of collisions that occur in one layer. At this level, collisions can only occur when two messages intersect and cause havoc on one or two hosts that are communicating.
- If the collisions of the session layer are intolerable, the control dialog will have another option: two-way alternate communication. Alternating two-way communication involves the use of a data token of the session layer, which allows each host to have its turn. This is similar to how a Token Ring layer 2 deals with the collisions of the first layer.
Separation of Dialogs
The separation of dialog is the beginning, end, and ordered management of communication. The main figure illustrates secondary synchronization. In the “Time Axis, t = checkpoint,” the session layer of host A sends a synchronization message to host B, at which time the two hosts will perform the following routine:
- Make a backup of specific files
- Save network settings
- Save clock settings
- Note the end node in the conversation
A primary synchronization involves more steps and talking back and forth than the one shown in this diagram. The checkpoint is similar to how a word processor on a stand-alone computer is interrupted for a moment when it performs an automatic save of the current document. However, these checkpoints are used, instead, to separate the parts of a session previously referred to as dialogs.