Improving scalability using asynchronous tasks
When building a desktop or mobile app, multiple tasks (and their underlying threads) can be used to improve responsiveness, because while one thread is busy with the task, another can handle interactions with the user.
Tasks and their threads can be useful on the server side too, especially with websites that work with files, or request data from a store or a web service that could take a while to respond. But they are detrimental to complex calculations that are CPU-bound, so leave these to be processed synchronously as normal.
When an HTTP request arrives at the web server, a thread from its pool is allocated to handle the request. But if that thread must wait for a resource, then it is blocked from handling any more incoming requests. If a website receives more simultaneous requests than it has threads in its pool, then some of those requests will respond with a server timeout error, 503 Service Unavailable.
The threads that are locked...
