We're really excited to be back from holidays with an awesome event up to come! We'll have 2 advanced Kubernetes talks. First of all, we'd like to thank our sponsor Octo technology (https://www.octo.com/en) for hosting this meetup.
Here's is the agenda:
Title : Kubernetes networking and the outside world
* Laurent Corbes After managing uptime and availability for large private cloud infrastructures, operating worldwide metro networks, and ensuring optimal QoS for a leading European CDN, Laurent is now VP of Infrastructure at Enix. Enix offers a wide range of infrastructure, consulting, and training services, from managed hosting to private cloud to bespoke container orchestration platforms.
Network communication within a Kubernetes cluster is usually considered to be fully solved: it is stable, and a vast number of CNI plugins are available. But what about exposing services to the outside world? Things are pretty straightforward when running Kubernetes through a cloud provider's managed offer; but on-premises deployment bring specific challenges. Let's talk about the available options, their pros/cons and how to implement them.
Title : Kubernetes in production: avoid being Hacker News' front page
* Yassine Tijani Yassine is a consultant for OCTO Technology. He is a Kubernetes org Member, where he co-leads several design proposals such as SchedulingPolicy, he is Member of sig-scheduling and sig-cloudprovider. He also works with customers to build production-grade kubernetes clusters and deliver solutions for Day-2 operations.
* Arnaud Mazin Arnaud is a consultant for OCTO Technology, deeply convinced by DevOps practices, He contributes to industrialize and automate the infratructures' and applications' lifecycle. He prefers to talk about practices and individuals rather than tools. His goal is to make the process of deploying to production as smooth as possible. He also gives trainings on Kubernetes, Docker and Ansible at OCTO Academy. Arnaud is one of the co-leads on the SchedulingPolicy proposal in the Kubernetes community.
Today, deploying a default kubernetes cluster isn’t much of a problem anymore: there are multiple tools at our disposition for easy deployment on a variety of environments, from bare kubeadm to kops, and cloud-specific products like acs-engine.
Nevertheless, there is a world between a “functional” cluster and a “production-ready” one, and several steps must be completed in order to offer a production grade cluster able to survive its initial deployment. Based on our field experiences and deployments made for numerous clients, we propose a non-exhaustive checklist of things to do to offer a robust, resilient, manageable and enterprise-ready kubernetes cluster: security, cluster design, disaster recovery, and many other concepts.
We're waiting for you to share a drink and have a chat.
We remind that the access to the venue is restricted to the registered attendees, so please don't forget to free your place if you can't make it :)