You are the Michael Jordan in System Design! Thanks man, I just completed the reading on this paper and nice to watch this video and refresh the topics.
Hi Jordan, I’ve been following your system design videos, and I find them incredibly insightful and well-researched. I’m curious to know how you research a topic before making a video. Do you watch other RUclips videos, read books or technical papers, or rely on any specific tools or resources? Understanding your research process would help me develop a framework that I can incorporate into my own learning journey.
Well for these most recent ones, I'm literally just reading a paper and summarizing it haha In prior ones, it's been some sort of synthesis of my intuition, articles, comments from users on prior versions of the videos, things like that! I just try to research things more when I don't understand them!
I told you don't fall in this trap, I've fallen and now I'm married xD. But the most important topic is now you can travel to Portugal and we can take a beer.
Hi, thanks for this videos, have really change the way o think things through. Anyhow, I was wondering, where do you get this info, and when from official papers, do you ever get the feeling/ certainty there may be pieces of the puzzle missing or even be misleading? is there a way to fact check somehow. For this social graph topic, is this syst. good enough? Or could’ve been done differently, do you know any smaller scale system that’s created its own Tao? Is Neo4j good or a scam? I heard its partners with Msft now, I don’t know what that could mean. Last question, is there a mother of all db tiers that stores all the history of facebook’s caches, reads and writes? I’m new to this tech level of things, I was just a converted ux and I just started having this need to go deeper and deeper. Thanks!
Hey! I'm just reading the papers is all I think that as far as pieces of the puzzle missing the best software is the one that gets the job done and in the case of TAO, it did for Facebook's specific use case. It doesn't necessarily mean it's the best tool for all graph data sets, but I do think that Neo4J at scale can be very tough to do given the interconnectedness of graphs in general, and their transactional support (IIRC). Point is, I'm hesitant to call anything better or worse than anything, because at the end of the day all of these systems are used in production and solved their business case!
the paper says "Neo4j is a popular open-source graph database that provides ACID semantics and the ability to shard data across several machines...To the best our knowledge, none of these systems scale to support Facebook’s workload"
You are the Michael Jordan in System Design! Thanks man, I just completed the reading on this paper and nice to watch this video and refresh the topics.
Hi Jordan,
I’ve been following your system design videos, and I find them incredibly insightful and well-researched. I’m curious to know how you research a topic before making a video.
Do you watch other RUclips videos, read books or technical papers, or rely on any specific tools or resources? Understanding your research process would help me develop a framework that I can incorporate into my own learning journey.
Well for these most recent ones, I'm literally just reading a paper and summarizing it haha
In prior ones, it's been some sort of synthesis of my intuition, articles, comments from users on prior versions of the videos, things like that! I just try to research things more when I don't understand them!
can't watch the whole video cause some uni BS courses consuming my time, but here to leave a like and a comment
Thanks for the content
I told you don't fall in this trap, I've fallen and now I'm married xD. But the most important topic is now you can travel to Portugal and we can take a beer.
Hi, thanks for this videos, have really change the way o think things through.
Anyhow, I was wondering, where do you get this info, and when from official papers, do you ever get the feeling/ certainty there may be pieces of the puzzle missing or even be misleading? is there a way to fact check somehow.
For this social graph topic, is this syst. good enough? Or could’ve been done differently, do you know any smaller scale system that’s created its own Tao? Is Neo4j good or a scam? I heard its partners with Msft now, I don’t know what that could mean.
Last question, is there a mother of all db tiers that stores all the history of facebook’s caches, reads and writes? I’m new to this tech level of things, I was just a converted ux and I just started having this need to go deeper and deeper. Thanks!
Hey! I'm just reading the papers is all
I think that as far as pieces of the puzzle missing the best software is the one that gets the job done and in the case of TAO, it did for Facebook's specific use case. It doesn't necessarily mean it's the best tool for all graph data sets, but I do think that Neo4J at scale can be very tough to do given the interconnectedness of graphs in general, and their transactional support (IIRC).
Point is, I'm hesitant to call anything better or worse than anything, because at the end of the day all of these systems are used in production and solved their business case!
the paper says "Neo4j is a popular open-source graph database that provides ACID semantics and the ability
to shard data across several machines...To the best our knowledge, none of these systems scale to support Facebook’s workload"
@ Yep!
Jordan I heard the bouncers kicked you out of TAO
Yeah I accidentally incited a riot when the ladies saw me
EDGING LMAO
Jordan, for the last time, you can't call your sock girlfriend judt because you use it and throw it in a dark room afterward... can you? 🤔🤨
Listen my sock has feelings too
Wow!