26 thoughts on “OpenStack Basics – Overview

  1. I just got done watching serveral OpenStack videos, and your video, up to
    now, was fare better than the rest.
    Thank you.

    So, from what I understand, the key components are:

    Newer software platform that is open and scalable.
    More access, or a open standard to the compute, storage and Networking.

    Are there any front runners in compute and storage, that have open up there
    side to work with OpenStack, and if so which ones?

    Duane Wellman Mattoon
    Senior FlexPod/Systems Architect 

  2. hi guys.. do i need to know unix or linux to install open-stack and admin
    it ? i am windows guy interested in cloud open-stack..please advise 

  3. Good overview of layer differences between openStack and more proprietary
    stacks. Was not aware before that openStack is targetting greenField for

  4. I do not think it is fair to say that OpenStack can’t be used for
    traditional application and meant only for new cloud aware application. 

  5. Thanks for all the feedback so far. We’re planning to make a new series of
    videos very soon, so we’ll have updates to include content for the “Folsom”
    release and looking forward to “Grizzly”.

  6. @deniskaran good catch. I don’t have time to change the video, but I’ll
    correct it hear in the comment. It should say, “applications designed for
    failures”, where the failure could either be at the HW or SW layer.

  7. Good Presentation in terms of the layered comparison between Closed-Stack
    and OpenStack. Wonder what and how Cisco will play a significant role with
    the solutions such as LineSider and etc. I’m a ex-Cisco, by the way.

  8. Swift’s purpose in life, and relationship to Nova isn’t characterized
    correctly here. Nova relies upon block storage for boot & persistence
    (although can also be used w/ shared storage outside of the context of
    nova-volume). Swift can be used as the backend for Glance, but images get
    copied out at instance creation. Swift is, IMO, more broadly useful as a
    standalone object store.

  9. That really depends on the size of the environment. If it’s a Private Cloud
    w/ specialized hardware (N+1), then their really isn’t a need to “Design
    for Failure” as the costs are much greater. If it’s a much larger
    deployment (~1K machines) using commodity parts and your running across
    multiple DC’s then “Design for Failure” becomes paramount.

Leave a Reply

Your email address will not be published. Required fields are marked *