Distributed Systems

Question 1

What is a distributed system?

Answer 1

A system that is distributed in nature, where components work together as one cohesive unit. It is fault tolerant and horizontally scalable.

Question 2

What are the main advantages of distributed systems?

Answer 2

1. Horizontal scalability, 2. High efficiency for given infra costs, 3. High availability

Question 3

What are the main disadvantages of distributed systems?

Answer 3

1. Increased complexity, 2. Requires expertise from multiple domains, 3. Data duplicacy, 4. Difficult data migrations, 5. Increased networking costs, 6. More difficult to secure, 7. Challenging deployments and troubleshooting

Question 4

What is reliability in distributed systems?

Answer 4

The ability of a system to perform its required functions under stated conditions for a specific period of time. It’s a measure of continuity of correct service.

Question 5

What is availability in distributed systems?

Answer 5

The proportion of time for which a system can perform its function as seen from a client’s perspective. It’s measured in percentage units with respect to time.

Question 6

What is scalability in distributed systems?

Answer 6

The property of a system to be able to meet increased load by adding proportional amount of resources without negatively impacting performance.

Question 7

What is fault tolerance in distributed systems?

Answer 7

The ability of a system to detect a fault and instantaneously switch to the redundant copy of the component with almost negligible downtime.

Question 8

What is consistency in distributed systems?

Answer 8

The ability of a system to maintain a single, up-to-date copy of the data, irrespective of how widely distributed it is.

Question 9

What does the CAP theorem state?

Answer 9

In a distributed system, one can only have either a consistent system or an available system in a partitioned network state.

Question 10

What is the PACELC theorem?

Answer 10

In case of network partitioning (P), choose between availability (A) and consistency (C); Else (E), choose between latency (L) and consistency (C).

Question 11

What are the ACID properties?

Answer 11

Atomicity, Consistency, Isolation, Durability

Question 12

What is atomicity in ACID properties?

Answer 12

A transaction must be treated as an atomic unit; either all of its operations are executed or none.

Question 13

What is consistency in ACID properties?

Answer 13

The database must remain in a consistent state after any transaction.

Question 14

What is isolation in ACID properties?

Answer 14

All transactions will be carried out and executed as if it is the only transaction in the system.

Question 15

What is durability in ACID properties?

Answer 15

The database should be durable enough to hold all its latest updates even if the system fails or restarts.

Question 16

What is a dirty read in concurrency control?

Answer 16

When one activity reads an uncommitted change made by another activity that is later rolled back.

Question 17

What is a non-repeatable read in concurrency control?

Answer 17

When one activity reads data, and another activity deletes that data before the first activity is done.

Question 18

What is a phantom read in concurrency control?

Answer 18

When one activity retrieves a set of data, and another activity inserts new data that would have met the first activity’s search criteria.

Question 19

What is pessimistic locking?

Answer 19

An approach where an entity is locked in the database for the entire time that it is in application memory.

Question 20

What is optimistic locking?

Answer 20

An approach that detects collisions when they occur and then resolves them, rather than trying to prevent them.

Question 21

What are the two main types of database storage engines?

Answer 21

1. B-Tree Based Engine, 2. Log Structured Merge (LSM) Tree Based Engine

Question 22

What are the main data replication strategies?

Answer 22

1. Log-Based Data Replication, 2. Full Table Data Replication, 3. Key-Based Incremental Data Replication

Question 23

What is algorithmic sharding?

Answer 23

Computing hash of the key in a record and computing modulo-n of that hash where n is the number of nodes.

Question 24

What is consistent hash sharding?

Answer 24

Uses consistent hashing technique to distribute data across nodes, with many more shards than actual number of nodes.

Question 25

What are the fallacies of distributed computing?

Answer 25

1. The network is reliable, 2. Latency is zero, 3. Bandwidth is infinite, 4. The network is secure, 5. Topology doesn’t change, 6. There is one administrator, 7. Transport cost is zero, 8. The network is homogeneous

Question 26

Name some common concerns in microservices architecture.

Answer 26

Configuration Management, Service Discovery, Load Balancing, API Gateway, Security, Centralized Logging and Metrics, Distributed Tracing, Resilience and Fault Tolerance, Autoscaling and Self-Healing

Question 27

What are some commonly used building blocks in a distributed system?

Answer 27

Load Balancer, Microservices, Caching Layers, Unique ID Generation Service, Scalable Databases, Schema Registry, Authentication Service, Service Discovery, API Gateway, Message Queues, Reverse Proxies, CDNs, Object Stores, Batch Jobs, Logging and Monitoring Dashboards

Question 28

What are some key design rules for building distributed systems?

Answer 28

1. Design for extra scale, 2. Avoid bleeding edge tech, 3. Optimize for important features, 4. Use generally available components, 5. Use caches extensively, 6. Use queues for transient data, 7. Avoid transactions, 8. Minimize IO, 9. Aim for idempotence, 10. Know your scale numbers