http
http copied to clipboard
Request context
As each request is handled in separate fiber it could be useful to store some information in fiber.storage (https://github.com/tarantool/tarantool/issues/3462)
- There is the stash. The README shows how to get a captured part of the URL and give a brief API description without explanatons. In fact, it is inspired by Mojo and works in the similar way: you can store a value here and load it back later during the request processing (from a template as well).
- You can just use
fiber.storage.
I'll consider this as the request to improve the documentation.
Yes, my initial thesis is not correct. All requests in the connection is handled by single fiber - it's an implementation detail of HTTP server. It's possible to use just fiber.storage.