Please don’t stop making videos!! I can’t believe you haven’t blown up but your videos are so good! The monad one was recommended to me on my home page and I love all your stuff!
I haven’t come across any other video on CAP that explains it as well as you did, and in simple terms - Thank you. Unicode is next in my watch list! Please keep creating more content like this.
Great job, as always. Also I really dig you topic choices. Stuff that everyone heard of at some point, but maybe was too lazy to research by themselves.
Thanks for the video. How can you have Consistency and Availability together without Partition tolerance?, basically if there is a Network Partition, you are not available anymore.
Just a thought...:What if we use some kind of a AI System to lead the parts together?->to keep availability at all times and Consistency trough the "right guess" of the AI System
Would an example of a hybrid solution during a partition be letting clients read data from the system, but not write data? That way, it can maintain consistency while providing "50%" availability.
_CAP_ is actually *PAC* "In case of *Partition* you can choose between *Availability* or *Consistency* but not both." *CAPELC* is the full reality: "In case of *Partition* you can choose between *Availability* or *Consistency* but not both. *Else* (when not partitioned), choose between *Latency* and *Consistency* "
a few days ago some of your videos popped up in my recommended list. thank you for your content! you offer very valuable information presented in a clear and concise way about software engineering, something i feel like comp sci students as myself can profit a lot from. Thank you, great job!
huh, i wonder how this is handled by big blockchains like bitcoin or ethereum? i'd guess the way proof of stake addresses it is different from proof of work
Your videos are great, with very clear explanations. Thanks a lot for sharing your work! I hope your channel takes off, you deserve way more subscribers.
I have one confusion: Lets say we have case for CP (Consistency and partition tolerence) Partition happens: and we want to keep consistency (not allow change in data) mean we lose availability as we are not able to write. Does this mean, its partition tolerant? How? If this does not allow write to the system, it is not usable any more? - A system that is active and running but not letting you to use it its not usable, then what is partition tolerence????
I would say "A system that is [...] not letting you to use it" is NOT "an active and running system" in the first place. Take RUclips's view counter for example: The part shown to us users is an A&P system: It's always available, it does synchronise / merge the number of VOD requests every X hours across the whole network, and MY region is not impacted by outages in OTHER regions (i.e. partition tolerant)... it just doesn't update in real-time and simply returns it's cached 301 counter, so it is inconsistent at the time of request. Remember that it's all about PRIORITIES in distributed systems. For C&P, I could sacrifice availability of a PART of my network to ensure it can be synchronised once it's online again with consistent, up-to-date data. As I understand it, Partition TOLERANCE is more about "How good can I deal with partitioning? To what extent do I ensure that every change will be respected / will be correctly merged and flushed to all machines eventually?" while a non-P system might decide to shut down completely or "fork" at some point. Consistency is more about "Do I ensure to only return a value that the whole network consistently agrees on, right at this moment?
Hi alex You are a clockwork i want say one question why you leave of roblox because you're account is dead and no one know you in roblox So do you want comeback in roblox
Please don’t stop making videos!! I can’t believe you haven’t blown up but your videos are so good! The monad one was recommended to me on my home page and I love all your stuff!
No new upload in 1.5 years 😭
he stopped right after this video... you jinxed it
What happened to you? No new videos in a year…
woooooo that is the most clear I have ever seen this. Never going to forget it now!
I haven’t come across any other video on CAP that explains it as well as you did, and in simple terms - Thank you. Unicode is next in my watch list! Please keep creating more content like this.
im always so gratefull for people like you, only reason i love learning , cz it makes sense when you guys explain
What other 18/30 minutes long videos can't properly explain to me, this 6 minute video did it with ease. Thank you very much.
Great explanation!!!
So far the best video I have seen to understand CAP theorem. Please do not stop making videos!
this is absolutely elite! teaching is a skill and this is it.
Great job, as always. Also I really dig you topic choices. Stuff that everyone heard of at some point, but maybe was too lazy to research by themselves.
please come back mr alex clockwork. i am 14 and these videos have intrigued me a lot and i can learn a lot from these.
Hey Alex, that was a great explanation. Thank you so much! Just discovered this channel. Love it!
This is superb, and the best explanation I've seen on this subject; thanks.
Thanks for the video.
How can you have Consistency and Availability together without Partition tolerance?, basically if there is a Network Partition, you are not available anymore.
Makes sense. Thank you very much ❤
I hope my man remembers his yt password 😔 Some of the best videos on the internet.
clockwork
Thanks alot for making this video. simplicity is beautiful
Your explanation is amazing
One of the best content in youtube at the moment. Please keep contributing to the learning community!
Just a thought...:What if we use some kind of a AI System to lead the parts together?->to keep availability at all times and Consistency trough the "right guess" of the AI System
Está super el video Alex
Whats the difference between CA and partition tolerance as a whole?, both returns an error, is it not the same?
amazing!
Would an example of a hybrid solution during a partition be letting clients read data from the system, but not write data? That way, it can maintain consistency while providing "50%" availability.
That's a perfect example 👌
4:24 beautifully put... the rewording makes it so much easier to understand.. it just cuts the clutter
_CAP_ is actually *PAC* "In case of *Partition* you can choose between *Availability* or *Consistency* but not both."
*CAPELC* is the full reality: "In case of *Partition* you can choose between *Availability* or *Consistency* but not both. *Else* (when not partitioned), choose between *Latency* and *Consistency* "
Stop receiving traffics means? system is available but not giving anything meaningful or not giving you are asking - am I right?
KATON GOKAYU NO JUTSU!
omg clockwork plz notic this omg ! !!!111
Easy to understand, but please remove the background music as it's distracting
Thank you. Please keep these Great videos coming.
Excellent.
Cool very well explained ..making such a complex concept look so simple ..
dont understand what your talking about but cool to see your still on the internet clockwork
clockwork?…
so thats why doos attacks are so bad, if one server is targeted it will create a partion which will mess up the traffic
What a program do you use to make these videos?
Excellent explanation Alex, thanks!
Would be better to consider two subsystems, read and write
Now explain Google Spanner or Amazon Aurora 😎
Clear explanation bro!
a few days ago some of your videos popped up in my recommended list. thank you for your content! you offer very valuable information presented in a clear and concise way about software engineering, something i feel like comp sci students as myself can profit a lot from. Thank you, great job!
Where does your radish pfp character come from?
Bruh, where have you disapppeared to?
You should make some new swe videos again. 👀
direct to the point, thanks !!
This is a great video, by the way.
Network partition is an inevitability
You can never have something which is CA
P is a natural event you either choose C or A sacrificing either one
But your system can be CA for indefinite time period until partitioning happens. That doesn't sound like "never".
Are you still making videos? Your videos are helpful please make some more.
Amazing video, best explanation i have seen yet
Been missing your videos...
Bro.... please come back..
great explanation👏
Bro, you still alive?
Nicely explained
Where you at?
clockwork
great job m8
don't quit!
super easy explanation, couldn't find such good anywhere else. Please continue making such videos, however, I would prefer no music in background :)
huh, i wonder how this is handled by big blockchains like bitcoin or ethereum? i'd guess the way proof of stake addresses it is different from proof of work
Nice
Please don’t stop making videos!! I can’t believe you haven’t blown up but your videos are so good!
I never understood CAP theorem specially Partition Tolerance. But your video was so to the point. Thanks a lot. Please bring more videos.
Excellent information explained in a concise, but understandable manner! Great infographics!
This is best youtube video that explained CAP in a simplest and quickest, and easiest way. Thank you so much.
One of the best explanations on the topic. I was looking for one for a friend who is learning and I wish I had this back then!
Thank you for the clear explanation!
last we heard from clockwork
Your videos are great, with very clear explanations. Thanks a lot for sharing your work! I hope your channel takes off, you deserve way more subscribers.
Neat.
Apart from these one comments, these videos have really helped me. Thanks alex.
Wow, amazingly explained!
YOU INTRODUCED ME TO WORLD ORDER THANK YOU BRO!!!
Awesome
At 5:26 (CP system), when partition appears and system responds with an error, it looses availability. Fair. But isn't partition tolerance lost too?
New videos when ?
so simple :) thank you
I hope you continue making videos. They're all really well done and I've seen channels like this blow up.
Hellow bro how we know which type is code
Amazing video!
What an explanation. 👌🏻
Great explanation in simple terms. Well done, Alex.
I have one confusion:
Lets say we have case for CP (Consistency and partition tolerence)
Partition happens: and we want to keep consistency (not allow change in data) mean we lose availability as we are not able to write.
Does this mean, its partition tolerant? How? If this does not allow write to the system, it is not usable any more? - A system that is active and running but not letting you to use it its not usable, then what is partition tolerence????
I have the same question. Let me know if you find an answer.
I would say "A system that is [...] not letting you to use it" is NOT "an active and running system" in the first place. Take RUclips's view counter for example: The part shown to us users is an A&P system: It's always available, it does synchronise / merge the number of VOD requests every X hours across the whole network, and MY region is not impacted by outages in OTHER regions (i.e. partition tolerant)... it just doesn't update in real-time and simply returns it's cached 301 counter, so it is inconsistent at the time of request. Remember that it's all about PRIORITIES in distributed systems. For C&P, I could sacrifice availability of a PART of my network to ensure it can be synchronised once it's online again with consistent, up-to-date data. As I understand it, Partition TOLERANCE is more about "How good can I deal with partitioning? To what extent do I ensure that every change will be respected / will be correctly merged and flushed to all machines eventually?" while a non-P system might decide to shut down completely or "fork" at some point. Consistency is more about "Do I ensure to only return a value that the whole network consistently agrees on, right at this moment?
ur videos are so awsome and useful clockwork :DDDDDD
I dont think that he wants to be called clockwork
@@SetUser OK man I'm just saying he is a legend
@@Razwizz yeah but the clockwork shades are cool
Grat video, but music is absolutely not necessary!
Hi clockwork
Hi alex You are a clockwork i want say one question why you leave of roblox because you're account is dead and no one know you in roblox So do you want comeback in roblox
ClockWork?
Help me
never expected this from clockwork
The DDIA book says this theorem is useless
Glad to hear other folks feel similarly :)
hi, are you clockwork?
so this is what clockworks is now......
sup alex do you have any social medias?