Big Misconceptions about Bare Metal, Virtual Machines, and Containers

preview_player
Показать описание

Checkout our bestselling System Design Interview books:

Other things we made:

ABOUT US:
Covering topics and trends in large-scale system design, from the authors of the best-selling System Design Interview series.
Рекомендации по теме
Комментарии
Автор

If you like what you see, SMASH subscribe. More videos are on the way ❤

ByteByteGo
Автор

This channel is PURE GOLD OMG!! The time, the love, the dedications spent in these videos, soft animations and simplifying everything is awesome, thanks for sharing your time❤Están espectaculares todos los videos!! :D

Capybara_Player_
Автор

As a total newbie I find this videos very easily digestible, and perfect to watch on 1.75x. This is some great work, thanks!

ketaminefairy
Автор

Thank you for always objectively expressing the pros and cons of various design approaches. As someone once said, all architectural decisions are trade-offs.

deadohiosky
Автор

Very good and clear explanation.

To be pedant, about "Bare metal is expensive hard to manage and hard to scale", it depends on many factors. In the most cases, it is undeniable true. But in some case e.g. you run app server on many machines and want to squeeze every single drop of your H/W, bare metal could be the cheapest. In some aspect, it is also possibly the easiest, since you cut the administration of between-layer like hypervisor or container management. In some aspect, not, like migration of app server to other machine.

kittipongpiyawanno
Автор

Good summary... while production compliant machines supporting popular bare metal hypervisors are pretty expensive, I've found that I've been able to install ESXi (my environment of choice) on a number of 'noncompliant' machines for test/eval... then go the expensive stuff for production (so folks interested in learning shouldn't be afraid to grab the free version of esxi and try installing it on one of their older machines) . Noisy neighbor can usually be managed with setup... without losing the ability to increase capital utilization by leveraging those machines for other tasks during less busy periods... and of course the ability of a production virtualization or containerization environment to optimize hardware utilization by moving workloads around (on the fly) REALLY kicks things up a notch.

kuhndj
Автор

Great quick overview about all three. Isolation, control & maximum hardware utilization are the true objectives of doing all this stuff.

viky
Автор

I've only watched like 3 videos but this is already one of my favorite channels! Thank you for providing this valuable knowledge with such great visualizations for free!!

r_mclovin
Автор

Smooth animations, direct language, and delivers 100% on its promise. Great content!

bernardomenescalferreirada
Автор

A note about 'bare metal' hypervisors, these are actually operating system with the minimum number of services and drivers needed to interact with ths hardware.
This is why any linux based operating system can be turned into a 'bare metal' hypervisor; the linux kernel itself is a 'bare metal' hypervisor thanks to it's kvm (kernel virtual machine) module

fuseteam
Автор

this is quickly becoming one of my fave YouTube channels. Your videos are amazing man, hats off

mario_luis_dev
Автор

Best ever explanation of VM v/s Container

shubhammguptaa
Автор

Excellent video, thanks for the upload. Coming from using computers in the 80's onwards, I immediately though "Bare Metal" = assembly language programming :) But yes, I can see that when talking about servers (or anything) we can, today, have different levels of abstraction. It's nice/calming to know (and expected) that Bare Metal here still = fastest...for exactly the same reasons - the more pathing you place between action initialisation and 'end-point' execution (i.e. machine code) then the slower the performance. Since machine code is the only language every CPU understands then, ultimately, that's what's running for every app/process.

ChrisM
Автор

System design can be understood pretty well if you've got a hang of the trade-offs that are made when choosing one implementation over the other!

aayushisingh
Автор

Definitely one of the best channels on YouTube. Great content

rva
Автор

Your videos are amazing! Your explanations are so easy to understand and fluent and the animations and the visual effects make them so much more interesting to watch. Overall thank you for these absolute gems and I hope to see your channel grow a lot.

saeethegreat
Автор

It was a refresher for me and thanks for explaining it very well. You spoke about edge computing in the end. And there is also a concept of Quantum computing. I wish if you can create similar videos to explain about both Edge computing and Quantum computing, I would really appreciate that :)

Vinod_Kumar
Автор

Thank you! Each one of your video is so high quality! Subscribed to see more to come!

solomonxie
Автор

Thank you for your succinct explanation of the differences between these computer architectures.

bobdinitto
Автор

I love this channel!
These are two superb professionals with excellent content!
Thank you for your work and content!

evdokiya