Can SCGI handle more than one request at a time?

Question

From my understanding of the SCGI specification, an SCGI server can only handle one request at a time. Is it true that SCGI servers can only handle one request at a time?

How do I work around this limitation, so that a web application served using SCGI is able to handle more than one request at a time?

I thought about starting more than one SCGI server. For example, if I start two SCGI servers listening at 127.0.0.1:8000 and 127.0.0.1:8001 respectively, I could configure the NGINX web server like this:

upstream myservers {
    server 127.0.0.1:8080;
    server 127.0.0.1:8001;
}

location / {
    include /etc/nginx/scgi_params;
    scgi_pass myservers;
}

Does this method actually allow my web application to serve more than one request at a time?

score 0 · Answer 1 · answered Feb 22 '22 at 13:08

It is not necessary to start more than one SCGI server to handle more than one request at a time, if the SCGI server itself is able to handle more than one request at a time.

For example, an SCGI server that uses the forking or pre-forking server model will be able to handle more than one request at a time. In the forking server model, the every time the server accepts a connection, it will fork a child to handle the connection, and then go back to accepting more connections. In the pre-forking server model, the server forks several children when the server starts. All of these children block on the accept() system call until a request arrives. When a request arrives, the operating system will unblock one of these accept() to let a child handle the request. While this child is handling the request, other children are still blocking on their accept(), ready to accept other requests.

If the SCGI server is multi threaded does it still need to fork ? It can also run an epoll based loop in a single thread and can handle multiple connection at the same time. — Neel Basu, Jul 16 '23 at 11:05

Can SCGI handle more than one request at a time?

1 Answers1