Distributed Cache System Design - Part I | What is Distributed Caching?

This video is the first part of my two part video course on designing distributed cache. In this video I am discussing:

00:00 - Introduction
00:45 - Purpose of Distributed Cache
02:40 - Read/Write Policies of Distributed Cache
11:00 - Cache Eviction Policies like LFU, LRU, MRU, FIFO etc.
14:00 - Functional Requirements for Distributed Cache System
16:35 - Non-Functional Requirements for Distributed Cache System

Check out our following articles:

#FAANG #Facebook #Google #Amazon #Apple #Microsoft #Uber #Netflix #Oracle #Lyft #SystemDesign #Interview #ComputerProgramming #DistributedCache #Cache

These two part videos discuss how a distributed cache like Memcache and Redis can be designed.

[ThinkSoftware]

Thanks for watching this video. Let me know your feedback. Thanks

[yassirhalim]

Very comprehensive as all you videos! Thanks a lot!

[sundeepjacob]

I went through your system design videos and I wanted to mention that the content is excellent! I hope you will make more videos, since these are tremendously useful for interview preparation. Thank you very much for helping people prepare

[Mohamed-ufjh]

Great job. It's really good thanks

[ibrahimshaikh]

Nice, liked ur video. plz make similar videos on load balancer. Thank you

[marklarization]

Hi, thanks for the great content! I have a question: why is strong consistency a requirement when you are using a cache aside pattern? With cache aside, the application cannot rely on the cache being consistent with the datastore, so what benefit does a strongly consistent cache offer to the application?

[vivekbhadoria]

Recently found your videos. Helpful - a big thanks. Question: The LRU and Cache Aside must be under non-functional requirements? Shouldn't they be covered under the consistency and cost? For a distributed cache design at the platform level - ideally, some of the eviction and read/write policies must be configurable.

[shailendraacharya]

really very nice video. I have one question in non functional requirements you mentioned that we need high availability and strong consistency. may be I am wrong but isn't it violation of CAP theorem? availability and consistency we can't guarantee both.

[khuranaishaan]

Hey! This was a quite insightful video. I got a question. If I use Heroku to create my servers, is it feasible to use HashMap for cache? I've seen my servers getting refreshed randomly, most likely every 24 hours. If that's the case, then all my data in the cache will be lost every 24 hours and it'll increase the app latency. What could be a workaround for that?

[lkank]

You are doing a great job. The content is really good and fresh. Just a suggestion, you can work on your presentation skill and stop reading during the talk.

[manu-qfoh]

hello sir ..
Could you please explain ... in the functional requirements, what is the difference between requirement 3. invalidate/delete and 5. eviction policy.
As per my understanding both are for data removal from cache.. then why you defined them as separate requirements?
Thanks.

[mithzeesh]

HI! Thanks for the amazing contnet.. You say that We cannot use Cache Aside and Read Through cannot be used if The Cache is getting updated.. Can you please explain in more detail WHY? I could not find anything about it in internet

[uXXair]

How can we have availability and consistency at the same time in a distributed system? Doesn’t that violates the CAP theorem?