Thank you for helping people understand that CAP is a framework for thinking about trade-offs in distributed systems, rather than a “law of nature”. It is often misused by companies trying to make black-and-white marketing statements about their products when the reality is more nuanced.
A very good very concise explanation of a very complex topic. I love the ATM example. It lets us understand the problem and even think of possible solutions like the "small withdrawals". If a node loses connectivity with 50% of the network, it must allow withdrawals up to the 50% of the amount registered just before the partition. Excellent!
can’t like or comment on these videos enough! every word and sentence is meticulously and concisely chosen, making these concepts clear and easy to understand. the visuals are a great cherry on top.
Super helpful video just like the book! I noticed that at the end of the video, there is thumbnail showing up on the left bottom, which blocks the content. If it's just talking kinda video, I understand, but for this kind of videos I found it's pretty vexing.
We should have a movement in system engineering to simplify the language that we use. The words network partition are ambiguous. A lot of people use the term network partitioning to describe a security strategy that limits blast radius to ensure availability and to stop lateral movement during an attack. The fact that the CAP theorem uses network partition to describe the networking being unavailable is contradictory. This is why everything needs extra effort to simplify understanding of people who are not engineers. We can, and should do better with use of plain and consistent language.
Thanks a lot! Neatly explained. Could you please also explain the interesting trade-offs to consider between latency and consistency in a similar manner? I am talking about what you mentioned at 5:14
Question: if P means continuing operating even in network partition and I choose CP then in case of partition how am I getting continuation of operation guarantee if operations are unavailabe (because of C)
Basically no one chooses this, as generally it would mean that the distributed system is not partition tolerant at all, which is the worst option. For instance by having a single pre-defined node taking over all responsibilities, or restarting all nodes and hoping the partition has gone away.
Thank you for helping people understand that CAP is a framework for thinking about trade-offs in distributed systems, rather than a “law of nature”. It is often misused by companies trying to make black-and-white marketing statements about their products when the reality is more nuanced.
Truest answer of all time “it depends”
This is now the best channel for learning System design and tech content. The videos are extremely high quality, concise and too the point.
A very good very concise explanation of a very complex topic.
I love the ATM example. It lets us understand the problem and even think of possible solutions like the "small withdrawals".
If a node loses connectivity with 50% of the network, it must allow withdrawals up to the 50% of the amount registered just before the partition.
Excellent!
can’t like or comment on these videos enough!
every word and sentence is meticulously and concisely chosen, making these concepts clear and easy to understand. the visuals are a great cherry on top.
Came to this video after hoping to many videos and blogs. So far the most crystal clear & best video.
ATMs and Twitter are 2 great examples to understand the trade off between consistency and availability in the presence of network partition.
Best explanation of CAP theorem I've ever seen
I learned CAP theorem from text books many times but never get it right. I would say the animations make it so much easier to understand finally
The best explanation ever! Thanks a lot!
Loving these videos. Especially when you relate them to the content in your books. The visual design is also crisp and top notch!
This channel is great looking forward pacelc theorem
Best explanation of CAP. Thanks 🙏
looking forward to the PACELC video!
Super helpful video just like the book! I noticed that at the end of the video, there is thumbnail showing up on the left bottom, which blocks the content. If it's just talking kinda video, I understand, but for this kind of videos I found it's pretty vexing.
One of the best explanations out there
final got a channel to start my system design journey
Very clear and just going to the point. Thank you!
The explanation is very easy to understand.
ByteByteGo's delivery of information is very efficient (Covers enough surface area, clear and concise).
the animation hits the right place.
First! ... on the server I am connected to :P
Very well explained. I watched 6 other videos and none were any good.
Thank you for the amazing explanation.
Could you help me understand how Google Cloud Spanner overcomes the CAP theorem?
Use only one ATM when the network is broken.
nicely
explained with examples
Love your videos! Thanks for sharing these pieces of knowledge :)
We should have a movement in system engineering to simplify the language that we use. The words network partition are ambiguous. A lot of people use the term network partitioning to describe a security strategy that limits blast radius to ensure availability and to stop lateral movement during an attack. The fact that the CAP theorem uses network partition to describe the networking being unavailable is contradictory. This is why everything needs extra effort to simplify understanding of people who are not engineers. We can, and should do better with use of plain and consistent language.
simple and excellent!
TYVM for your videos Sir.
Very well explained… thank you
Thank you
Thanks for this nice and accurate explanation.
Thanks a lot! Neatly explained. Could you please also explain the interesting trade-offs to consider between latency and consistency in a similar manner? I am talking about what you mentioned at 5:14
Thank you for this video!
Excellent tutorial on cap theorem 👍👍
Thank You. Well Explained.
great job..thank you soo much
the best explanation
Awesome!! ty
I think a good follow up video would cover data structures and algorithms like CRDTs and raft.
Please make videos on AWS Services. Event Driven Architectures, Monolith Architectures, Lambda Functions, EC2 instances, S3, IAM Service.
OMG, it's so clear
awesome, ty!
Easy to follow, thanks, Alex. I don't see PACELC in Volume 1. Is it described in Vol 2?
awesome content
I love the animations, are they 100% in house or is there a certain program you use?
Question: if P means continuing operating even in network partition and I choose CP then in case of partition how am I getting continuation of operation guarantee if operations are unavailabe (because of C)
Please make a video about PACELC
great video, thanks for sharing
Does any one know software author is using to create these clear presentations
Hi, can you make a video explaining what a pipeline is?
Thank u for video. U explain CP,AP. But when happening CA 🤔
Basically no one chooses this, as generally it would mean that the distributed system is not partition tolerant at all, which is the worst option. For instance by having a single pre-defined node taking over all responsibilities, or restarting all nodes and hoping the partition has gone away.
🙌🏻
Best video
What software is this presentation made of??
Substantive content.. a bit busy.. with all the movement on screen though
support
Cryptocurrency, yes or no?
Ha ha
Good videos, good explanation, good graphics and so bad accent. Man, please do something with voice.
CAP seems no longer be true with Aws Aurora, Google Spanner etc.
new topic to learn
Thanks😊
Why not? It's got 99.99% availability but doesn't mention anything about consistency.
🧢 theorem. 😜
Fucking genious explanation