Pages

Thursday, August 4, 2016

Fedora Flock - 2016 - Day 3

Third day!  Before I get started on my session logging today, check out the picture of all the attendees at flock that we took last night before the cruise of Krakow.


Today we started with lightening talks for an hour.  I was second up and presented +OpenShift on +Fedora Project.  That was my first time presenting a lightening talk and my first time attending other lightening talks.  I really like the format for both.  You'd be surprised at how much material you can cover in 5 minutes.

Today is also hack session day.  The sessions I am attending are "Building a Fedora Containers Library", "OpenShift on Fedora", and "Fedora PRD Workshop".  The sessions are two hours each.

+Josh Berkus kicked off "Building a Fedora Containers Library" with a slide that had instructions to git clone the lab material.  That's the proper way to start a workshop :). Josh walked us through building a +PostgreSQL image step by step with lots of best practices discussed along the way.  This session was particularly insightful because Josh is well... extremely knowledable on PostgreSQL.  That knowledge coupled with his Docker chops translated into an outstanding session. Great hack session.

Next was "OpenShift on Fedora", a hack session led by Maciej Szulik.  The material for the lab is located here.  We started out by leveraging vagrant to spin up an environment that we could issue an "oc cluster up" in, which spins up everything you need to get started.  The lab consisted of deploying pods, exploring pods, services, replication controllers, etc.  Maciej did a great job explaining some concepts in OpenShift that I wasn't really getting.  Such as deployment configs, image streams, horizontal scaling.  I didn't quite finish with the lab but the good thing is you can take it with you in the Vagrant box and the material on github.  Great hack session.


Final session of the day was the PRD session.  This is where different members of the Fedora workgroups came together to discuss strategy and how to move forward.  Very interesting session to see how the sausage is made.  We got through the Server and Workstation groups and then transitioned into Cloud / Atomic.  Unfortunately we didn't have to much time to get into the details - spent time focusing on vision and mission statement.  I'll record the tactical things I'd like to see done here that I think we can actually accomplish:

  • Drive Fedora Atomic into cloud providers: GCE, Azure, AWS
    • Think marketplace quickstarts here
    • Think ansible playbook driven deploys here
  • Two options for Cloud providers
    • All in one for developers / ops to kick the tires on
    • Production quality one for ops to kick the tires on 
  • Fedora based CDK
  • Perhaps start work on documentation / demos for ^
This was another useful session.  I'd like to see the Cloud workgroup dedicate some time to ironing out the rest of the PRD.

Today ends with a stint a a local brewery.  Another great day of Flock in the books.

No comments:

Post a Comment