System design is the most challenging and exciting type of tech...

Here is a complete list of topics that will help you get ready:
{0/18} ↓
Machines in distributed systems communicate with each another over a network.
What to know:
- basic understanding of low-level networking
- protocols (IP, TCP, UDP, DNS, and HTTP)
Information storage is a fundamental component of every distributed system.
What to know:
- volatile vs. nonvolatile memory
- the concept of a database
Most distributed systems need to be highly available.
What to know:
- how to measure a system's availability
- how to increase a system's availability
- how to use redundancy
Caches store responses to network requests or computationally expensive operation results.
What to know:
- cache hit and miss terminology
- cache eviction policies
- content delivery networks
These are key parameters to evaluate the performance of a distributed system.
What to know:
- latencies of the most common operations
- back-of-the-envelope calculations
Proxies are intermediary servers used in every system.
What to know:
- forward and reverse proxies
- use of reverse proxies as load balancers
- load balancer's server selection strategies
Hashing is helpful for the elastic scaling of cache servers and data partitioning.
What to know:
- consistent hashing
- rendezvous hashing
These are structured databases storing data in tabular format and supporting SQL queries.
What to know:
- indexing
- ACID transactions
- strong vs. eventual consistency
Key-value databases (i.e., Redis, Zookeeper ) are often used for caching and configuration.
Other databases to know:
- blob storage (S3)
- time-series databases
- graph database (Neo4j)
- spatial databases and quadtrees
These are standard techniques to increase availability and performance.
What to know:
- how to duplicate data on multiple servers to increase redundancy
- how to divide data across multiple servers to increase throughput
This is how a cluster of servers selects a leader responsible for all the primary operations.
What to know:
- what is a consensus algorithm
- how Paxos or Raft works
These are the most common techniques to obtain data from a server.
What to know:
- how to fetch data at regular intervals (polling)
- how to get a continuous data feed (streaming)
Every system needs to measure performance and troubleshoot issues.
What to know:
- how to collect and log events information
- how to get visibility of the system key metrics
- how to aggregate human readable metrics
This is a widely used messaging pattern.
What to know:
- how pub-sub works
- what are idempotent operations
- popular frameworks like Apache Kafka
These are machines splitting a workload between them to complete it the fastest way.
What to know:
- what is a gossip protocol
- use cases of such networks
Limiting the number of requests sent to or received by a system is essential.
What to know:
- how rate limiting can prevent DDoS attacks
- how to implement rate-limiting strategies (i.e., with Redis)
This is a popular framework for processing large distributed datasets efficiently and in a fault-tolerant way.
What to know:
- what is a distributed file system
- popular distributed file systems implementations (i.e., HDFS)
What to know:
- the basics of Web API design
- the concept of CRUD operations
If you liked it, I'd be grateful if you'd:
• like or retweet the 1st tweet
• follow @Franc0Fernand0 for more distributed system content
• subscribe my newsletter polymathicengineer .com (link in bio)
Your support encourages me to keep writing!
There I will provide detailed explanations and additional resources.
The next issue is about caching.
The link for subscribing is on my bio: polymathicengineer .com