Upgrading your Cluster with Zero Downtime

In this episode of Kubernetes Best Practices, Sandeep Dinesh shows how to upgrade your Kubernetes cluster without disrupting your applications and services on Google Kubernetes Engine.
See the associated article here → bit.ly/2LOPLf2
Take a look at more Kubernetes Best Practices videos here → goo.gl/xTzAGd
Subscribe to the Cloud channel → goo.gl/S0AS51
Regional Clusters → bit.ly/2LOPhFK
Node Auto-Upgrades → bit.ly/2LS5mdU
Node Pools → bit.ly/2LP7nHO
Google Kubernetes Engine → bit.ly/2LS5gTA
#KubernetesBestPractices

[@IceW4ve]

Very good explanation, as always. Thanks!

[@IsozwWz]

Cordon not only makes a node nonscheduled for new pods, but also the Load balancer no longer redirect traffic to perfectly healthy pods on a cordoned node.

[@vishwanathk6706]

When you cordon and drain the node one at a time, there is a chance that the drained pods can get scheduled on old node pool nodes and not necessarily on new node pool nodes.

[@stouneua]

Should this recommendation be deprecated in favor of using deployments and Horizontal Pod Autoscaler?

[@jweyrich]

Great lesson! Thank you for this!

[@jweyrich]

Actually, thank you for the entire series/playlist!

[@cyberstair]

Please upload more videos on this topics

[@AnilKumar-de4lw]

If we were to update Node-selector in every Name-space deployment yaml file, we could not proceed in this manner. The new node pool name must be updated in the NodeSelector yaml file section, and then the instructions must be followed. Provide the solution when node-selector is enabled in every yaml file. How to efficiently complete the task of upgrading.

[@bhavikattipraveen3030]

nice explanation sandeep and i would like to know any official online course in advance level ,already i have some idea till intermediate level practiced too from offical docs and i wanna do more complex projects or tasks under your training ?????? .... urgently from your end Sandeep