Aug 28, 2018
What does it take to support Kubernetes for other users? Kenneth Massada, a lead for GKE support at Google Cloud, tells Craig and Adam his story.
Do you have something cool to share? Some questions? Let us know:
- web: kubernetespodcast.com
- mail: kubernetespodcast@google.com
- twitter: @kubernetespod
Chatter
- Adam lives in Seattle, which is on fire
- Craig
baked some tasty cookies
- Using this recipe
- But not using Vegemite, British Marmite or
New
Zealand Marmite, which are three totally separate things. Only
one of which is nice.
- Hint: it’s the last one
News of the week
- 2018 Kubernetes Steering Committee Elections
- Binary Authorization on Google Kubernetes Engine
- kube-hunter from Aqua Security
- Kubernetes issues and solutions from Alexander Lukyanchenko at Avito
- Cilium 1.2 released
- James Lee’s blogs on Kubernetes networking
- Amazon EKS supports GPU-Enabled EC2 instances
Links from the interview
- etcd is hard:
- Configuration flags
- OpenAI suggestions on scaling Kubernetes to 2,500 nodes includes a separate events database
- Kubernetes docs on configuring and upgrading etcd
- Tina and Fred from Google SRE also discussed etcd on Episode 9
- (Or use GKE, where we do it all for you)
- Other hard concepts:
- apiVersion: is hard
- spec: is hard
- Liveliness and readiness probes - don’t make them the same!
- Joe Beda thinks of YAML as machine code in Episode 12
- What would Ken like to see changed in Kubernetes?
- Affinity and anti-affinity rules and topology keys
- Kenneth Massada on Twitter