Systems Design Elements

Question 1

Metadata database

Answer 1

A separate database that only handles metadata

Question 2

Chunking functionality

Answer 2

An app/service that breaks file data in to chunks and rebuilds it

Question 3

Listener/watcher functionality

Answer 3

Monitor a particular element for changes (e.g. via long polling) and notify another target element

Question 4

Indexing functionality

Answer 4

Process changes in one file and updates the metadata database with information

Question 5

Message queueing service (and 2 main components)

Answer 5

Asynchronous, loosely coupled communication between distributed system components. 2 main elements:

• Request Queue - shared between all clients
• Response Queues - unique to each client

Question 6

Data deduplication handling (and 2 ways)

Answer 6

Eliminating duplicate copies of data by calculating hashes of data chunks and comparing to existing chunks
2 ways:
* post process: chunks are stored first then a cleanup process removes duplicates later
* in-line duplication: hashes are calculated in real time

Question 7

Throttling (3 types)

Answer 7

• hard throttling
• soft throttling
• elastic/dynamic throttling

Question 8

Hard throttling

Answer 8

of API requests cannot exceed a hard limit

Question 9

Soft throttling

Answer 9

of API requests cannot exceed a % factor of the overall rate limit

Question 10

Elastic or dynamic throttling

Answer 10

of requests can exceed threshold if system has resources available

Question 11

Rate limiting

Answer 11

definition of the rate and speed at which clients can access APIs

Question 12

Rate limiting - fixed window algorithm

Answer 12

requests are limited within fixed time windows regardless of when requests are made

Question 13

Rate limiting - rolling window algorithm

Answer 13

requests are limited within rolling time windows based on when requests are made