Kubernetes Cluster Autoscaler Vs HPA | Amazon EKS Cluster Autoscaler Demo

In this video, we discuss kubernetes cluster autoscaler vs horizontal pod autoscaler (HPA). We do a full end to end working demo of cluster autoscaler in Amazon EKS.
Reference Link: docs.aws.amazon.com/eks/lates...
Connect with me in LinkedIN: / rajdeep-sa-at-aws
Follow me on twitter: / cloudwithraj
DEFINITELY DO NOT CLICK: / @cloudwithraj

[@cloudwithraj]

My Kubernetes on EKS Course is out now in Udemy!! If interested, please check out discounted ($9.99) link - www.udemy.com/course/rocking-kubernetes-with-amazon-eks-fargate-and-devops/?couponCode=GETSKILLSNOTCORONA

[@vishakarudhra8665]

I am really thankful you put up this video. Because without lectures of this kind, a person like me with limited comprehensibility of the language used in documentations could never actually understand the essence ever.

[@cloudwithraj]

Yw! And yes, sometimes k8s documentation is filled with jargons

[@MichaelRicksAherne]

Thank you so much for this demo. I would have been able to figure this out eventually, but you've saved me hours and hours of time.

[@nforlife]

Great Demo clean, clear and easy to understand!!!

[@cloudwithraj]

Glad it was helpful!

[@kayedmasarweh7422]

Great video Raj .. Clear and easy to understand

[@cloudwithraj]

Kayed brother, this means a lot coming from you. Thanks for watching man, I am glad you liked it.

[@venugopalreddypagidi8880]

Thank you so much Raj

[@arunmanglick]

This is wonderful

[@big0561]

Fatafati chilo !! Your explanation was just right on the money.

[@cloudwithraj]

Onek dhonyobad Tirthankar bhai :).

[@bhupathibrahmaiah7651]

Thank you for making this video it really helped to understand and implement CA in EKS. If possible can you make a video on EKS VPC CNI with demo

[@nagarajunagari2266]

Thanks a lot Bro for valuble information... All the best..

[@cloudwithraj]

So nice of you. Thanks for watching!

[@Asgardinho]

this videos explains better than anything else I have find, could you do one explaining EKS fargate with autoscaling and another one explaining EFS connection to EKS

[@cloudwithraj]

Thank you so much for the kind words! I am so glad you found this video useful.

[@MhdBasharDas]

Thank you for the useful video, but how the cluster will auto-scale down even there is no traffic (it's not dependent on CPU usage as you mentioned).

[@Sravankumar-ip9yo]

Hey raj can you please explain how we can attach new worker node to the existing cluster

[@heohful]

Hey Raj, great video! 1 quick question - After replicas are increased in the deployment file and it's redeployed, the nodes are auto created but the pods don't associate with the new node created even though the nodeSelector is right. It constantly shows in pending state. Any thoughts here?

[@himansusekharmishra5114]

Hi Thank you for this awesome video. I just want to know that I have an existing cluster with separate nodes like one node is in the public subnet another is in private subnet and I want to make that particular private node as a autoscaling node and secondly how can be the number of nodes increase(cluster autoscaling behave) without adding manual replica set?

[@user-xz7gx9kr5z]

Thank you so much

[@devopspanther]

Thanks for the demo. How come there is no service account and oidc used for the CA ?

[@shrey2539]

Hello, May i know why the IAM permissions and tags were skipped? Is it because kubectl automatically takes care of that part?

[@George-mk7lp]

what strategy EKS use when there is no pod hpa ? How does it define resource requests and limits

[@vijaykumary5427]

What about if my pod cpu is more than my request and less than the limit set... Will the autoscaler increase the node count?

[@pidugureddy]

Thanks for explanation... I saw one more nice video on AWS EKS Autoscaling with CA and HPA on this channel. #techtter

[@cloudwithraj]

Thank you so much for the kind words! I am so glad you found this video useful.

[@kaustubhmishra3973]

The eksctl command to create a cluster is so easy. I had to create the full infrastructure using cloudformation. Can you please forward the link to these commands. I am not sure these simple commands are shown in aws docs anymore.

[@ajaykharat8832]

eksctl create cluster --name test --region us-east-1 --version 1.15 --managed --asg-access Error: invalid version, 1.15 is no longer supported, supported values: 1.22, 1.23, 1.24, 1.25, 1.26, 1.27

[@joanale609]

that document was updated anyone have this downloaded?

[@ECOM-EXPLORER]

Your course is very nice. I have a question like in aws we have feature like we can scale based on the queue length. is there a way we can scale in Kubernetes on basis of queue length?

[@cloudwithraj]

Yes, you can. You need to grab custom metrics (queue depth) outside of standard metrics (CPU, Memory). Check out keda.sh/

[@juanwigg]

Thank you for this awesome video! Cluster autoscaler works fine along with HPA?

[@cloudwithraj]

Yes it does! thanks for watching.

[@sanketgund4924]

thanks. keep it up. want a video for HPA

[@cloudwithraj]

Already have one with demo :), here you go - kzbin.info/www/bejne/aXPRo4uXpb5qsNE

[@xamashee]

Very comprehensive and simplified. Just would like to ask how to disable scaling of worker node in sandbox environment?

[@cloudwithraj]

Don't use HPA or Cluster Autoscaler and just use fixed value in replicaset to keep it at fixed size. or you can use HPA to utilize the nodes you have and don't set up cluster autoscaler

[@xamashee]

@@cloudwithraj sounds practical. Will discuss with team internally! cheers!

[@GopalRoy-nn6ft]

Where is cluster Overprovisioner?

[@MrKumargopalan]

kindly do one training on daemonset Raj

[@cloudwithraj]

Will definitely keep in mind Ramkumar, thanks for watching!