Thread by @svpino

Caching is one of those techniques that can completely change how your application (website) feels when used.

💼 As a developer, you want to understand caching and incorporate it into your projects whenever possible.

This is a thread 🧵about caching.

⬇️

📘 In short: caching is fast storage.

You want to keep often-used data in a place where you can access it much faster than its source.

For example, instead of downloading data from the server over and over again, you can store it in memory and access it from there.

👇

To keep the cache from growing indefinitely, you want to maintain the most relevant entries cached and evict everything else.

There are multiple strategies to accomplish this. Each one makes a different decision about which items to remove and which to keep.

👇

Here are 8 popular caching strategies:

▫️First-in, first-out (FIFO): Evicts the oldest entry

▫️Last-in, first-out (LIFO): Evicts the newest entry

▫️Least Recently Used (LRU): Evicts the least recently used entry

👇

▫️Most Recently Used (MRU): Evicts the most recently used entry

▫️Random replacement (RR): Evicts a random entry

▫️Least-frequently used (LFU): Evicts the entry that has been used the least

👇

▫️Least frequent recently used (LFRU): Combines LRU with LFU

▫️Time aware least recently used (TLRU): Combines LRU with expiration for each entry

👇

💼 Deciding when to invalidate entries in your cache is a hard problem to solve.

For example: if you are going to expire entries after a certain time, how much time is adequate? What happens if the content is updated before?

A lot of creativity needed here!

👇

There’s a famous saying:

There are only two hard things in Computer Science: cache invalidation and naming things. — Phil Karlton

So yeah, getting caching right is not easy.

But it’s well worth it!

👇

Finally, for the Python 🐍 developers out there, here are some resources about caching that you should check out:

* functors.lru_cache decorator
* cachetools library
* Beaker library

JavaScript developers, what libraries do you use to cache stuff?