Communicate
The interactive parameter transmission of stand-alone applications in the same address space is obtained by a function call or global variable access. Inter-component communication between different address spaces in the network environment is completed by the following communication mechanisms:
l Message transfer and shared memory
Message Transfer: Application developers typically need to define the format and capacity of these messages, and the agreement adopters will follow them exchange information. The protocol itself also defines the number of participants in the one-time message interaction, such as point-to-point, broadcast, and multi-point transmission, but also define how message interactive participants start, perform and end a message. At present, many popular intermediate architectures use messaging-based communication modes, such as CORBA.
Shared memory is divided into two, including local shared memory and distributed shared memory
Ø Locally shared memory: A nice choice for multiple threads or processes in a single machine. But this has brought harm to the applicability of these applications, and their implementation has become careful.
l Distributed Shared Memory: This is a higher level programming abstraction, she provides the application with network distributed virtual memory. This way expands the concept of virtual memory, and the system provides synchronization updates to each process and different network contacts. This model often maintains coordination and synchronization between readers in the bottom layer.
Server architecture
Single service server only provides a service.
Multi-service server integrates a series of single service servers to a management unit with the following advantages:
Ø Reduce system resource overhead
Ø Simplify server development implementation, reuse effective code
Ø Allow external services to upgrade the service without modifying other existing code and interrupt server
Ø Strengthen monitoring management of network services. This can be done by some network application configuration management programs
Disposable server and persistent server
The disposable server will only start when the request arrives, and ends after the request is triggered.
Lasting server
In addition to the life cycle of the service, they usually start when the system starts or is first called by the customer.
