Giters

brlala / System-Design-Basic

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

5 Tips for System Design Interview

  1. Do not get into detail too fast.
  2. Avoid silver bullet(fix mindset)
  3. K-I-S-S Keep it small and simple
  4. Justify why you want to make a simple choice
  5. Aware of current technologies

Techonologies to use

  1. Load Balancing(Nginx)

    • For Webserver

  2. Caching(Memcached, Redis, CDN to cache static assets file)

    • makes around the world fast, PULL/PUSH to CDN Method

    • For image server

    • For database server(reads)

    • Least Recently Used, Least Frequently Used, First In First Out eviction policy

  3. Slave Master Replication

    • for database server(reads)

  4. Databases + Indexes

    • RDBMS - MSSQL/MYSQL For relationship and speed of access
    • NoSql scalability - MongoDB, HBase(supports small write and read)
    • Cassandra NoSql - Masterless, low downtime
    • HBase - Minimum five datanodes and one namenode, high maintenance, require the use of JRuby shell and interdependency like ZooKeeper, high learning curve

  5. Database Sharding

    • For database server(write)
    • Horizontal (consistent hashing or hash on keys)/Vertical Sharding (each server has different tables)

  6. Load Balancer

    • Determine what algo to use like Round Robin, Weighted Round Robin, Consistent Hashing

  7. Pagination

Basic Steps for Systems Design

  1. Determine the functional requirements
  2. Determine the non-functional requirements (high availability, data should not be lost)
    • data consistency (seeing same message for all devices)
    • availability (replication)
    • reliability (persistent)
    • latency
  3. Determine extended requirements(accessible through REST API's, analytics, push events etc)
  4. Limit on content/text (how many data/string can a user push)
  5. Traffic estimates
    • Users
    • Files per user
    • Storage per file
  6. Storage estimates
    • How long to keep the data
    • char(1 byte), int(4 byte), date(4 byte)
  7. Bandwidth estimates
  8. Memory estimates
  9. System APIs (Parameters/Function calls)
  10. Database design (scheme)
    • determine how much each row would be and
    • if there is relation
    • read/write heavy

Limitations and solutions

  1. Maximum 500 connections for webservers(it will block)
    • split read and write servers and have dedicated services for each other to ensure the system does not block, also allows scaling and optimizing
  2. Reliability and redundancy
    • multiple replicas and at least 1 replica in case primary has an issue
    • must be a number of 3 if using master-slave approach
  3. Data sharding
    • Determine number of shards to use and use UserID % 10
    • For hot users, we can partition based on photoID, by having a dedicated separate database instance to generate auto-incrementing IDs. To also solve the issue of Single Point of Failure, we can have two KGS with one generating odd and the other generating even and put a round robin infront.
  4. Sending Data to Users
    • Pull, clients can pull contents from the server on a regular basis:
      • New data will not be shown until client issue a pull request
      • Most of the time pull request will result in an empty response if there is no new(difference) data
    • Push, maintaining a long poll request with server to receive updates
      • Server has to push update frequently
    • Hybrid:
      • Separate pool for different users
  5. Reading data
    • We can maintain latency by pushing contents to cache servers/CDNs
    • Use Cache between Client-Server, Server-Database
    • In addition, for hot users

Summary

My framework for system design: 1st clarify the domain to design data model, from there design the CRUD API, from there figure out data flow, R/W frequency, persistency requirements which naturally leads to operations/scaling. In short:

  1. Data Model
  2. API
  3. Data flow
  4. Scaling

Watch In Order

  1. https://www.youtube.com/watch?v=vvhC64hQZMk Designing WhatsApp
  • image
  1. https://www.youtube.com/watch?v=oUJbuFMyBDk Queue
  2. https://www.youtube.com/watch?v=FMhbR_kQeHw Publisher Subscriber Model, drawbacks is consistency
  3. https://www.youtube.com/watch?v=GeGxgmPTe4c Distributed consensus
  • image
  1. https://www.youtube.com/watch?v=xrizarXJgC8 Avoid cascading failures in a distributed system image
  • Caching
  • Gradual deployment
  • Coupling(etc. Save authentication in server memory and assume it works for the next few hours)
  1. https://www.youtube.com/watch?v=K0Ta65OqQkY What is Load Balancing?

About