A distributed system is a collection of autonomous computing elements that appears to its users as a single coherent system.
Types of Distributed System
- High performance distributed computing: Cluster, Grid, Cloud, Fog
- Examples: Singapore Supercomputer, Amazon EC2 Cloud Service
- Distributed information systems (分布式信息系统): Ledgers, Databases, Storages
- Examples: Public Blockchain Network, Apache Cassandra Systems
- Pervasive systems (普适系统): Ubiquitous computing, Mobile computing
- AI 解释: 是一种将计算能力无缝集成到环境中的技术理念，旨在让人们在任何时间、任何地点、以任何方式获取和处理信息，而无需过多关注计算机本身的存在。
- Examples: Self-Driving Vehicular Network, Wearable Devices
分布式系统的特征
Autonomous 的内涵
- 每一个节点都独立运作
- 每一个节点拥有自己独特的 notion of time，因而不存在 global clock
Coherent 的内涵
- 对用户隐藏内部的复杂度，transparency
分布式系统的设计目标
Resource sharing
- Enable collaborative access and data exchange.
Distribution Transparency
Distributed systems aim at making the distribution of processes and resources (objects) transparent (invisible) to the end users and their applications.
| Term | Hides … |
|---|---|
| Access | … how objects are accessed |
| Location | … where objects are located |
| Migration | … whether objects is relocated |
| Replication | … that objects are replicated |
| Concurrency | … concurrent access/sharing |
| Failure | … failure & recovery of objects |
Openness
An open distributed system offers components that
Scalability
Scalability has become a fundamental goal for modern distributed systems, especially due to the heterogeneity of nodes, resources and components.
| Aspect | Explanation |
|---|---|
| Size | 用户、资源增多的情况下，处理计算与存储需求 |
| Geometry | 用户分布各个地区时，处理通信需求 |
| Administration | 处理跨独立组织的系统管理问题 |