Distributed systems must overcome a fixed set of recurring hardships beyond what single-machine programs face: heterogeneity, openness, security, scalability, failure handling, concurrency, and transparency.
Key facts
The “eight fallacies of distributed computing”: network is reliable / zero latency / infinite bandwidth / secure / topology stable / one admin / zero transport cost / homogeneous — all false.
Distributed Systems may use completely different hardware and software. These listed above are examples of differring hardware/software:
Networks
Computer Hardware
Operating Systems
Programming Languages
Implementation by different developers
These problems have some solutions such as:
Using standard protocols
Using agreed upon message formats and data types
Adhering to an API
Middleware
Portable code (code that is not tighly coupled to specific hardware)
Middleware
Middleware is a software layer between the distributed application and the operating system that:
provides a programming abstraction and
masks the heterogenity of underlying platform
Middleware Models
Distributed File Systems
Remote Procedure Call (RPC)
Remote Method Invocation (RMI)
Distributed Documents
Distributed Databases
Openness
Openness refers to the ability to extend the system in different ways by adding hardware or software resources. Following are some approaches to address openness:
Publishing key interfaces
Allowing a uniform communication mechanism to communicate over the published interfaces.
Ensuring all implementations adhere to the published standards.
Security
There are three aspects of security:
Confidentiality
Integrity
Availability
Security Mechanisms:
Encryption
Authentication
Authorization
Scalability
A system should be able to handle growth in users.
Scalability Challenges
Cost of physical resources
Controlling performance loss
Resources should not run out
Avoiding performance bottlenecks
Failure Handling
Detecting : some types of failures can be detected and some are hard to be certain about.
Masking : some failures that have been detected can be hidden or made less severe;
Tolerating : it is sometimes impractical to try and handle every failure that occurs
Recovery : some failures can recover, for example a rollback mechanism
Redundancy : some failures can be made to tolerate failure using redundant components (eg multiple services that provide the same service(failover))
Concurrency
Multiple clients can access the same resource at the same time, in some cases for updates.
One approach to handling concrrency is making access sequential - slows down the system.
Semaphores supported by the operating system is a well accepted mechanism to handle concurrency.
Transparency
Hiding certain aspects from the user and application programmer