All You Need to Know in 12 Minutes: Pods' Requests and Limits in Kubernetes

In this deep dive, learn all about requests and limits in Kubernetes. Discover what they are, why they are important, and how to configure them for optimal performance. With this 13-minute video, you'll have a solid understanding of how to manage resource allocation in your Kubernetes cluster. Watch now to optimize your knowledge of this essential Kubernetes feature.
Twitter - / the_good_guym
LinkedIn - / guy-menahem
YT - / @the_good_guy
#deepdive #kubernetes #pods #request #limits #capacity #resourcemanagement #container #k8s #oom #oomkill #oomkilled #cpu #throttling #node #nodes #management #application

[@orencrazy]

Great content! Thanks for clearing these for us.

[@viniciusfialho2691]

Awesome content! Thanks for sharing!!!

[@the_good_guy]

Thanks for watching!

[@estebanlopez5372]

How you explain is awesome! Thanks a lot! You're really a good guy :)

[@the_good_guy]

Thank you!!! 🙏

[@birajsilwal]

very helpful content and explanation

[@the_good_guy]

Thank you! 🙏

[@emredemir6637]

Thanks for the video and the tip about metrics. Ah also, which Catan game is it you added the video (2.01-2.07)?

[@the_good_guy]

I have no idea 😅😅

[@the_good_guy]

I have no idea 😅😅

[@mexicanmomo]

Yeah.. great content . Now I understand why my application runs smoothly in a single pod , but slows down or stuck when I scale it to 4 pods..either memory or cpu at fault. It's not memory probably, it's cpu..

[@the_good_guy]

Thank you! I'm glad I could help 😃

[@pratapgavali4786]

Great information, thanks for this,
I've a question, currently there are three pods getting created for my website, and each pod is consuming max to max 4% of my request cpu, so can I reduce request and limit of the cpu so it'll not waste rest of the cpu, like should I reduce it by 80%-85% so that after it'll utilize given cpu only and will get cpu usage numbers to 70 to 80% of request. Please tell me thanks

[@the_good_guy]

I think that you should have a little bit of headspace and reserved resources. Reducing the requests and limits sounds great but I would do that gradually and monitor for CPU throttling at the same time to make sure nothing is impacted by the change.

[@pratapgavali4786]

Thanks for the confirmation, I'll do the same. 😊

[@YuchenDeng-1007]

Amazing video! Thanks!
I have a question, I wrote the deployment file with limits, but why the limit didn't work
deployment file is like this:
spec:
containers:
- name: rt-tests
image: yuchen1007/stress-ng:v.1
resources:
limits:
memory: "1200Mi"
cpu: "1"
requests:
memory: "1200Mi"
cpu: "1"

[@the_good_guy]

Thank you!
What's the error you got?

[@YuchenDeng-1007]

@@the_good_guy I set the cpu limit but the limit doesn't work

[@the_good_guy]

How do you know it doesn't work? It's very hard to identify that your application has a CPU limit without having a metrics dashboard in place

[@YuchenDeng-1007]

@@the_good_guy l use htop to check cpu limit

[@YuchenDeng-1007]

@@the_good_guy l use htop to check the cpu limit