Which two guarantees can a distributed system NOT provide simultaneously according to the CAP theorem?

Consistency and Partition Tolerance

Availability and Partition Tolerance

In the context of the CAP theorem, what does 'Partition Tolerance' mean?

The system can handle network failures or partitions between nodes.

All nodes in the system must have the same data.

The system guarantees that a response will be returned for any request.

The system can scale to handle a large number of requests.

Stop serving requests to avoid inconsistent data.

Continue serving requests from all nodes, potentially leading to stale data.

Serve requests only from nodes that can communicate.

Replicate data aggressively across all available nodes.

A NoSQL database designed for high availability, like Cassandra.

A relational database with strong ACID guarantees.

A single-node application server.

A system requiring immediate data synchronization across all clients.

Consistency vs. Latency (when the system is NOT partitioned)

Partition Tolerance over Latency

Latency over Partition Tolerance

Partition Tolerance over Availability

System designers must understand their application's requirements and make conscious trade-offs based on these theorems.

It's always possible to achieve all three properties (C, A, P) if the network is fast enough.

The theorems are theoretical and have little impact on real-world distributed system design.

The PACELC theorem makes the CAP theorem obsolete.

Focusing solely on Availability is the best strategy for all distributed systems.

CAP and PACELC Theorem Quizzes for System Design — Free MCQ Practice Test | testpaper.space