Fast... But Wrong? Meet Cache Invalidation
This is Part 7 of my "From One User to One Million" series, where we'll build an understanding of System Design by following a simple application as it grows from a single user to millions. Instead of memorising technologies, we'll learn why they exist by solving real problems as they appear. Last time, we ended on a question that sounded simple but isn't. Aisha updated her profile picture. Her…
Section 1: When Cached Data Lies
Cache invalidation poses a hard problem in software systems, differing from complex algorithmic challenges. The severity of stale data's impact depends on the cached content. For instance, caching trending articles shows mild annoyance when outdated, but product pricing cached inaccurately can lead to frustrated users and support issues.
Similarly, cached user permissions may grant unauthorized access, raising security concerns. The core issue is that any time data resides in multiple locations, discrepancies arise given enough time and writes. A famous quote highlights that cache invalidation and naming things are the two hardest problems in computer science.
Section 2: The First Instinct: Delete It When It Changes
The apparent solution to cache invalidation is straightforward: when the underlying data changes, delete the corresponding cache entry. This approach, known as Cache Aside, works as follows: during reads, the cache is checked first. If the data is found, it's returned; otherwise, the database is queried, the data is stored in the cache, and then returned.
On write operations, the database is updated, and the cache entry is deleted. Using Aisha's profile picture update as an example, the process involves updating the profile in the database and deleting the cache entry for 'profile:aisha'. When the next request arrives for Aisha's profile, the cache miss triggers a database fetch for the updated data, which is then cached for subsequent requests.
While this method ensures data consistency, it necessitates knowing which cache entries to delete upon every data write. This can become cumbersome in complex systems with multiple database tables and application components composing data. As the number of data sources and operations increases, determining which cache keys to invalidate becomes challenging.
The quote about cache invalidation being one of the hardest problems in computer science underscores the difficulty of maintaining cache consistency amidst frequent data updates.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
