The Operating System | OS | Part 5
Architecture / Comparison between monolithic kernel or microkernel
An operating system is typically organized into distinct layers. The top layer is the programming interface with application software (which includes software utilities included with the operating system).
At the center includes one or more layers that contain the main components of the operating system such as programs for file systems and networking, memory management (see above), the pilot, the scheduler, the interrupt handler.
The lower layer called HAL (Hardware Abstraction Layer, short: HAL), is responsible for hiding the physical characteristics.
The Core / Kernel Operating System
The core or kernel as it is well-known is a single memory space, which is placed in all or part of the operating system. In common parlance, the term kernel refers to the location and all of its programs, and forms the heart of the rigid operating system. The content of the nucleus can not be inadvertently altered by software applications – a modification would crash the computer.
Monolithic Kernel
The architecture is said to be monolithic kernel where all the programs of the operating system (especially the drivers, programs that address the network and file system) resides in kernel space. In this architecture each time a function of the programming interface causes a switching of context and the transition kernel mode.
Linux, FreeBSD, OpenVMS and Solaris operating systems are said to be monolithic kernel.
Micro-Kernel
The architecture is called micro-kernel (also called microkernel) when the nucleus contains the bare minimum, i.e., the scheduler and the program that simulates virtual memory, and the vast majority of programs are outside: drivers, programs that deal with file systems or graphical user interface and application software.
In this architecture the many programs that are outside the core are isolated from each other, are implemented as application software – in competition – and use the core functionality for exchanging messages. The use of functions of the programming interface results in the emission of new messages.
Minix, BeOS, Mac OS X or QNX operating systems are examples of microkernel.
Hybrid Kernel
Various operating systems have an architecture that has some characteristics of micro-nuclei and also nuclei monolithic. This architecture is called core-core hybrid macro or micro-kernel changed.
In this architecture, all the programs of the operating system reside in the nucleus, like a monolithic kernel. However many programs operating system are executed in competition such as software applications, like the micro-kernel architecture (kernel processes).
For kernel processes, the scheduler performs context switching between processes in the same manner as between application software, these processes use the functions of the scheduler for exchanging messages and use functions programming interface causes the transmission of messages.
Windows NT and NetWare are operating systems to hybrid kernel.
Exo-Kernel
In the exo-kernel architecture (Greek exos = off), there is no single location, and the operating system is composed of programs and software libraries construction identical to those software applications.
The use of functions of the programming interface causes the execution of instructions in a software library operating system without requiring a context switch. A software library is used for transmitting and receiving messages between processes, the scheduler instructions are included in this library.
Study: From Wikipedia, the free encyclopedia. The text is available under the Creative Commons.
