There was so much happening on the opening day of OpenStack that this blog will be a bit longer than I usually do for events but I want to make sure you can get all the information.

Highlights:

  • Over 1,300 attendees were announced this morning in a standing room only keynote room. Even more impressive was the typical large conference three screen setup found at traditional conferences.

Some notes from the session:

  • Crowbar - build a complete infrastructure from scratch for multi-node system (OpenStack, Hadoop) based off the DevOps model (coding your operations)
  • BarClamp - individual items that are loaded during the install process (e.g. BIOS, DNS, MySQL, CEPH, Force10, Nagios, Ganglia)
  • Install the cloud in layers building upon itself (see image below)
  • Crowbar 2.0 - (release later this year) remove Crowbar specifics from Chef cookbooks, heterogeneous OS, network install, pull from source

  • Dell Session: Getting from Folsom to Grizzly: A DevOps Upgrade Pattern with Greg Althaus, Principal Engineer

Some notes from the session:

  • Problem - how to handle the complexity of upgrading OpenStack releases on a regular basis?
  • Offered three possible ideas for managing the upgrade process in a DevOps manner (#1 On the Fly / #2 Split-Migrate-Replace / #3 See Image)

Look for even more information tomorrow from various sessions, an interview with Joe Arnold from SwiftStack, and other interesting items I discover during the day. For live tweets during the day watch #OpenStack and/or my feed at @SpectorAtDell.