What you need to know to successfully run databases in production on Kubernetes

The Kubernetes community has made significant progress when it comes to easily deploying stateful services like databases. But what happens on Day 2? Day 2 operations are all about managing an application when things go wrong: nodes fail, networks are partitioned, a CVE comes out requiring a new version of Kubernetes to be deployed and all running apps upgraded. This talk will provide a real-world view into the challenges of running stateful apps like databases in production and look at some solutions using Portworx PX-Enterprise.

In this session, Portworx Engineer Michael Ferranti will discuss and demo:

--How to architect an application to take advantage of the self-healing provided by Kubernetes

--How to respond to common Day 2 operational scenarios like node failure, disk out of space, and restore from snapshots

--How to conduct a blue-green deployment of a new version of Kubernetes when you need to upgrade your cluster and keep your databases running

For more training like this, visit:

[adamjones]

I'm half way through, and I'm very appreciative to have an industry veteran break down the state of things and provide so much insight. My biggest request, if anyone reads these, is PLEEEEASE fix the audio glitches! It makes it very hard to follow along when the keyword in a sentence gets dropped! I understand it was originally a webcast and his connection might be inconsistent, but requesting them (in the future) to simultaneously record an audio file and then matching it up afterwards would turn this from an artifact of a webinar into an actual piece of educational material. It was certainly well planned and delivered (unlike some material on youtube with umms, uhhhs, unsupported advice and sometime incorrect information...).

[vasvd]

How database used in production with kubernetes