Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Micro-Blogging for The Enterprise (MongoDB)

Micro-Blogging for The Enterprise (MongoDB)

Transition from relational structure to MongoDB

Amit Kumar

April 26, 2011
Tweet

More Decks by Amit Kumar

Other Decks in Technology

Transcript

  1. Amit Kumar Consultant and Rubyist TATA Consultancy Services Limited twitter.com/toamit

    github.com/toamitkumar Who Am I Sunday, September 16, 12
  2. AGENDA Micro-Blogging for “The Enterprise” Relational DB Structure Performance issues

    Caching? MongoDB Hybrid Approach Sunday, September 16, 12
  3. Message threads (Facebook like experience) Activities of follower list Two

    timelines Enterprise Demands Sunday, September 16, 12
  4. Message threads (Facebook like experience) Activities of follower list Two

    timelines Individual Enterprise Demands Sunday, September 16, 12
  5. Message threads (Facebook like experience) Activities of follower list Two

    timelines Individual Public Enterprise Demands Sunday, September 16, 12
  6. Message threads (Facebook like experience) Activities of follower list Two

    timelines Individual Public Filtering messages (role/office etc) Enterprise Demands Sunday, September 16, 12
  7. Performance Benchmark SQL query benchmarks 10,000,000 messages 200,000,000 comments 50,000,000

    message likes 2 4 8 10 12 0 750 1500 2250 3000 Response ms in millions Sunday, September 16, 12
  8. Caching Solution Like Break point in Architecture No Query Interface

    No clustering support Sunday, September 16, 12
  9. Caching Solution Like Break point in Architecture No Query Interface

    No clustering support More messy code Sunday, September 16, 12
  10. A Document Based DB? Storing them as relational structure Converting

    them back to documents Sunday, September 16, 12
  11. A Document Based DB? Storing them as relational structure Converting

    them back to documents Store them as “Activity Documents” Sunday, September 16, 12
  12. Performance Benchmark 2 4 8 10 12 0 750 1500

    2250 3000 SQL QUERY vs MongoDB Response in ms in millions SQL MongoDB Sunday, September 16, 12