Peer-to-Peer Computing in Distributed Systems

A distributed system is a collection of independent computers that appear as a single system to users.
File-server provides an interface for clients to store and retrieve files.

Peer-to-peer computing is another model of a distributed system where nodes are considered peers.
In P2P computing, nodes can act as clients, servers, or both.
To join a P2P network, a node registers its service with a central lookup service or broadcasts a request for service and responds to requests for service via a discovery protocol.
Examples of P2P computing include Napster and Gnutella.

The web has become ubiquitous, with PCs being the most prevalent devices.
More devices are becoming networked to allow web access.
A new category of devices, load balancers, manages web traffic among similar servers.

Open-source operating systems make the source code available in addition to the binary code.
This movement is counter to the copy protection and Digital Rights Management (DRM) movement.
The Free Software Foundation (FSF) started the open-source movement, which includes the GNU Public License (GPL).
Examples of open-source operating systems include GNU/Linux, BSD UNIX, and Sun Solaris.

A process is a program in execution and needs resources to accomplish its task.
Resources include CPU, memory, I/O, files, and initialization data.
Process termination requires the reclaim of any reusable resources.
A single-threaded process has one program counter specifying the location of the next instruction to execute.
A multi-threaded process has one program counter per thread.
A system typically has many processes, both user and operating system processes, running concurrently on one or more CPUs.

The operating system is responsible for creating and deleting both user and system processes.
The operating system suspends and resumes processes.
The operating system provides mechanisms for process synchronization and communication.
The operating system provides mechanisms for deadlock handling.

All data must be in memory before and after processing.
All instructions must be in memory to execute.
Memory management determines what is in memory when to optimize CPU utilization and computer response to users.
Memory management activities include keeping track of which parts of memory are currently being used and by whom, deciding which processes and data to move into and out of memory, and allocating and deallocating memory space as needed.

The operating system provides a uniform, logical view of information storage, abstracting physical properties to a logical storage unit (file).
Each medium is controlled by a device (e.g., disk drive, tape drive).
File-system management involves creating and deleting files and directories, primitives to manipulate files and directories, mapping files onto secondary storage, and backing up files onto stable (non-volatile) storage media.
Mass-storage management involves free-space management, storage allocation, and disk scheduling.

The I/O subsystem is responsible for memory management of I/O, including buffering, caching, and spooling.
The I/O subsystem provides a general device-driver interface and drivers for specific hardware devices.

Protection is any mechanism for controlling access of processes or users to resources defined by the operating system.
Security is the defense of the system against internal and external attacks.
Systems generally first distinguish among users to determine who can do what.
User identities (user IDs, security IDs) include a name and associated number, one per user.
User ID is then associated with all files, processes of that user to determine access control.

Traditional computer environments are blurring over time.
Office environments use PCs connected to a network, terminals attached to mainframe or minicomputers providing batch and timesharing.
Home networks use firewalled, networked systems.
Client-server computing uses smart PCs as clients, requesting services from servers.
Clustered systems are multiple systems working together, sharing storage via a storage-area network (SAN), providing high-availability service.

Multiprogramming is needed for efficiency, organizing jobs (code and data) so the CPU always has one to execute.
Timesharing (multitasking) is a logical extension in which the CPU switches jobs so frequently that users can interact with each job while it is running.
Response time should be less than 1 second.
Each user has at least one program executing in memory (process).
If several jobs are ready to run at the same time, CPU scheduling is required.
If processes don't fit in memory, swapping moves them in and out to run.
Virtual memory allows execution of processes not completely in memory.

The operating system is interrupt-driven by hardware.
Software errors or requests create exceptions or traps.
Dual-mode operation allows the OS to protect itself and other system components.
Mode bit is provided by hardware, allowing the system to distinguish between user code and kernel code.
System call changes mode to kernel, and return from call resets it to user.