KubeCon London 2016 - Day 1

By Mike Metral
Thu, Mar 10, 2016 using tags KubeCon , Community , Enterprise , OpenStack , Monitoring

   Vote on Hacker News  

Hello from London!

Today was the first day of KubeCon Europe in London, England, and the excitement & energy at this event has been nothing short of electrifying!

Lets review a few summaries & notes on a subset of the talks given today.

Around 500 folks from around the globe gathered to review the latest & upcoming features of the Kubernetes project, as well as, how the community is converging on various efforts to strengthen & solidify the foundation, plus, unveil how they’re using Kubernetes internally.


Day 1 Opening Keynote: Kubernetes Update

  • Talk Abstract
  • Summary
    • Kubernetes is a framework for building distributed systems
      • It is not a PaaS
    • Tip: Make sure you use version numbers with your images
      • The latest tag is not a version to deploy
    • If you’re new to Kubernetes, an easy, out of the box route to play with the project is via kubectl
      • It allows you to on-ramp with using the project, rather than jumping straight into dealing with Kubernetes manifests
    • Pods Refresher
      • Logical application includes:
        • 1+ conatiners and Volumes
        • Shared namespaces
        • One IP per Pod model
    • Tip: Do not try to convert the world into Kubernetes right off-the-bat - start small
    • Note: ReplicaSets replaced ReplicationControllers
    • ConfigMaps (new as of v1.2)
      • New Feature: A mechanism to create loose descriptions for configurations either through environmental variables or files, and it notifies the application to restart itself so that it can reference the updated settings
    • Ingress offers:
      • Integration with load balancers
      • Manage SSL/TLS for the first time
      • Note: This is still a work in progress
    • Manifest features to use/integrate into your apps
      • lifecycle: steps to run on container quit/termination
      • livenessProbe: enable a check system for app readiness
    • Scaling with kubectl is too imperative
      • Use Deployments instead to be more state-aware & manage your apps server-side, rather than something like kubectl scale client-side

Kubernetes State of the Union

  • Talk Abstract
  • Summary
    • New releases are occuring, on average, every 3 months
    • Velocity
      • 5k commits in 1.2
      • +50% Unique contributors
      • 1200+ External projects based on k8s
    • Vision
      • Give everyone the power to run agile, reliable, distributed systems at scale
    • Kubernetes v1.2 Release
      • New UI
      • Improved scale - 400% increase in number of nodes used
      • Simplified application deployments
      • Automated cluster management
      • Third party extensibility
      • Available by end of March
    • The Future
      • Kubernetes 1.3 has already begun
        • Legacy app support (aka “Pet Set”)
        • Cluster Federation (aka Ubernetes)
        • Scale++
        • In-cluster IAM
        • Cluster autoscaling
        • Scheduled Job object (i.e. cron jobs)
        • Public Cloud Dashboard (w/ nightly runs to indicate Cloud performance metrics and integration+support)
        • Open for proposals now
    • Cloud Native Computing Foundation (CNCF)
      • The entire Kubernetes code base has been donated to the CNCF
      • The CNCF has an aim to be an open foundation, such as the Apache Foundation, to support containers and the ecosystem around it
        • Kubernetes is one of the key projects in the CNCF
      • The CNCF is part of the Linux Foundation

What is OpenStack’s role in a Kubernetes world?

  • Talk Abstract
  • Summary
    • What are some benefits of running Kubernetes on OpenStack?
      • OpenStack gives you options for containers
        • i.e. If you decide to run Mesos, or Swarm in addition to Kubernetes, you have that ability
      • OpenStack is natively multi-tenant
      • You get more than just infrastrucutre, you also get a toolkit
        • i.e. Swift, Trove, Sahara etc.
      • OpenStack is future-proof
        • It’s ready for whatever is next because it not only serves the needs of its users today, but it also addresses the needs they’ll have tomorrow
        • New projects are always being started with an aim at tackling existing problems & adding new functionality
          • i.e. Kuryr for advanced networking, potential Unikernel support in the future etc.

Hybrid Apps: Orchestrating Cloud-Native and Traditional Application Architectures

  • Talk Abstract
  • Summary
    • Apps can be cloud native, enterprise IT is not
    • Options for adopting Kubernetes for Enterprise
      1. Decomposability requires conversion into miroservices - this just isn’t possible for most enterprises
      2. Leveraging CI/CD tooling - this is an ok model and works for most enterprises, but only solves the “push” problem in deployments
      3. Leverage a platform that understands the mechanics of “mixed era” apps - Kubernetes becomes a big part of this mixed solution, and Apprenda aims to help enterprises solve most of “the rest”
    • Kubernetes support for Windows is coming, and Apprenda plans on being a big part of this effort

Monitoring Microservices: Docker, Kubernetes, and GKE Visibility at Scale

  • Talk Abstract
  • Summary
    • Key features any container monitoring solution should have:
      • High level exploration and resource usage by Services, Pods, etc.
      • Low level visibility
      • Avoid container instrumentation
      • Low overhead
      • Smart configuration
    • Simple options for monitoring
      • Kubernetes UI
        • Great starting point to see resource utilization and give a logical view of your cluster
      • Configure Heapster + Grafana
    • Low level visibility is hard and doesn’t play well with containers (see below)
    • Solutions for containers
      • Sysdig
        • Aims to capture events
        • Filter, drilldown, explore & aggregate information
        • Is container native
        • Open Source
        • Supports Kubernetes v1.2

   Vote on Hacker News