Go to the U of M home page
Showing posts with label caching. Show all posts
Showing posts with label caching. Show all posts

Wednesday, August 21, 2024

Tony Tech Terms: Caching

Caching is a process that temporarily stores frequently accessed data to make retrieval faster and more efficient. Instead of fetching information from the original, often slower source each time a user or system requests it, the system can quickly access the data from the cache. This speeds up data delivery, improving performance and responsiveness across applications.

There are several ways to implement caching. In our systems, we use Oracle to cache data in the form of "materialized views" that refresh on a schedule. We also cache data as JSON-formatted text, which is effective for slower, once-per-day operations, like pulling data from Ad Astra for roomsearch.umn.edu. Sometimes, the browser caches information in a "session," while other times, the server temporarily stores data in memory.

Caching does have trade-offs. If the cache isn’t updated properly, we may encounter stale data, and, as you’re likely aware, clearing your browser cache is sometimes necessary. Despite these drawbacks, caching is usually well worth the benefits it provides in enhancing performance.