« Feedblendr Architecture - Using EC2 to Scale | Main | Scaling Early Stage Startups »
Tuesday
Oct302007

Database parallelism choices greatly impact scalability

Sam Madden in the The Database Column blog covers some database architectures. Quick summary:
  • Shared-memory systems don't scale well as the shared bus becomes the bottleneck
  • Shared-disk systems don't scale well either
  • Shared-nothing scales the best
  • Reader Comments (2)

    Praise for shared-nothing is only meaningful if there's also discussion of the cost. Is it harder to build and debug, and if so, how much? What's the comparative cost of coding errors that survive into production deployments? If the answers are "no" and "the same", then it's a clear win, but I doubt they are.

    December 31, 1999 | Unregistered CommenterGreg Wilson

    > Praise for shared-nothing is only meaningful if there's also discussion of the cost.

    Good point. Shared nothing isn't easy, but I've found the setup and maintenance of the other architectures pretty hard as well. Maybe that's just me though.

    December 31, 1999 | Unregistered CommenterTodd Hoff

    PostPost a New Comment

    Enter your information below to add a new comment.
    Author Email (optional):
    Author URL (optional):
    Post:
     
    Some HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <code> <em> <i> <strike> <strong>