-
Notifications
You must be signed in to change notification settings - Fork 6
Hitbox Server
Andrey Ermilov edited this page Nov 28, 2021
·
9 revisions
Hitbox Server is a caching reverse proxy for HTTP.
Adding the caching layer to the project causes a number of problems:
- increasing the codebase and complexity of project support
- sometimes you have to write a lot of code because caching libraries require integration with the framework you use, and this does not always go out of the box
- when you scale your project horizontally there appear questions about the effectiveness of caching use
The solution to the above problems can be creating a sidecar application that will take care of all the caching work, leaving the application to take care of the business logic.
- It is possible to put Hitbox Server between users and application (one of the most usual cases) and cache the answer to user's requests
- You can put Hitbox Server between the application and a third-party service or database and thus cache hits from the application
- CDN
- integration with most popular databases and the ability to use any of them as a cache storage
- store cache on disk (be persistent)
- easy installation and setup for basic cases
- incredible customization options for specific load profiles (stale cache mechanics, cache locks for dogpile effect preventions)
- great support for standalone and for distributed applications
- offload cache (how to describe it?)
- https://varnish-cache.org/
- https://github.com/Trendyol/sidecache
- https://doc.traefik.io/traefik-enterprise/middlewares/http-cache/
- https://www.krakend.io/docs/backends/caching/
- https://www.envoyproxy.io/docs/envoy/latest/start/sandboxes/cache
- https://www.haproxy.com/blog/accelerate-your-apis-by-using-the-haproxy-cache/