Introduction to Scaling Memcache

Key-value stores and big data

When designing large-scale social media websites, we need to understand that there are huge volumes of data and that a number of requests are being made. The large scale of data on such platforms entails very intense load and spontaneous I/O requests, which might not be optimized using prefetchingPrefetching loads data into the cache before the user requests it..

This is where caching in memory can be used to allow for linear scaling. A key-value store is one way to cache data. Key-value stores can run on network-enabled systems that share information with other machines without requiring specialized hardware. This allows architects to scale their key-value stores by adding more front-end servers. One such key-value store is Memcached.

Note: We'll call the distributed key-value store as MemcacheMemcache without the 'd'. and the source code or the running binary as Memcached.

Why scale Memcache?

Memcache allows developers to add caching to their applications while having simple operations like set, get, and ...