Читать книгу Cloud Computing Solutions - Группа авторов - Страница 79

Cloud computing involves request-response communication between the clients and stateless server. In the case of stateless server, a client doesn’t need to first establish a communication to the server; instead, the server views a client request as an autonomous transaction and responds to the client.

Cloud computing uses a stateless server due to some of the following obvious reasons:

 If the server node fails due to some technical faults or natural disaster, it will create a considerable overhead because of maintenance of the state of all the connections.

 If a stateless server goes down, the clients won’t be affected because no data or no information from the client’s side will be lost.

 The clients need not be concerned about all the states of the server; if a client gets a response to a request, that indicates the server is running and quite capable of handling a request.

 A stateless server is more advantageous than a connection-oriented service. In the connection-oriented service, the states of all the connections have to be stored and that will take up space in the memory. So, the stateless server is also memory efficient.

The common example of a stateless server is a web server. The web server need not keep the history and states of the past interactions with client; it just responds to the HTTP (Hypertext Transfer Protocol) request made by the clients. HTTP is basically a request-response application protocol which is used to communicate between the web servers and the clients, i.e., a web browser. Web browsers are also stateless; neither the clients nor the web browsers need to reserve the state of the server; they just send the requests and wait for the reply from the web server.

The following subsections deal with the service-oriented architecture (SOA), the need for SOA in cloud computing, its characteristics, and combining the strategy of SOA and cloud architecture.