SECTION A
(Short answers – explained clearly)
1. Why would you design a system as a distributed system? List some advantages.
A system is designed as a distributed system to allow multiple computers to work together and share resources over a network. Distributed systems improve performance by allowing tasks to run in parallel on different machines. They also provide better reliability, because failure of one node does not stop the entire system. Scalability is another advantage, as new machines can be added easily to handle increased load. Resource sharing and improved availability make distributed systems suitable for large-scale applications.
2. What is a proxy? Give an example where a proxy can be used.
A proxy is an intermediate object that represents another object located in a different address space, usually on a remote machine. It acts on behalf of the real object and handles communication between the client and the server. For example, in a web browser, a proxy server is used to fetch web pages from the internet for users, improving performance and security by caching responses.
SECTION B
(Medium answers – conceptual explanation)
3. What is distributed transparency? Explain different types of distributed transparency.
Distributed transparency refers to hiding the complexity of a distributed system from users and programmers so that it appears as a single unified system. Different types of transparency include access transparency, which hides differences in data access methods; location transparency, which hides the physical location of resources; and replication transparency, which hides the presence of multiple copies of data. These transparencies make distributed systems easier to use and manage.
4. What is Mutual Exclusion? Why is it more complex in distributed systems?
Mutual exclusion ensures that only one process at a time can access a shared resource. In distributed systems, mutual exclusion is more complex because there is no shared memory and no global clock. Processes must coordinate through message passing, which introduces delays and failures. Network latency and node crashes make it harder to ensure fairness and correctness compared to a single-computer system.
SECTION C
(Long answers – detailed explanation)
5. Describe Three-Phase Commit Protocol. How is it different from Two-Phase Commit?
The Three-Phase Commit (3PC) protocol is a distributed transaction protocol designed to reduce blocking problems found in the Two-Phase Commit (2PC) protocol. In 3PC, an extra phase called the “prepare to commit” phase is added between the voting and commit phases. This ensures that no participant is left in an uncertain state if the coordinator fails. Unlike 2PC, 3PC avoids indefinite blocking and improves fault tolerance in distributed transactions.
6. Explain naming in distributed systems. What is flat naming and structured naming?
Naming in distributed systems refers to the way resources such as files, services, or objects are identified. Flat naming uses simple, unique identifiers with no structure, such as random numbers or hashes. Structured naming organizes names in a hierarchical manner, such as directory-based file systems or URLs. Structured naming is easier for humans to understand and manage, while flat naming is efficient for system-level identification.
