I am hosting, on Windows Server 2022, a web application that makes regular ajax calls to a PHP backend.
The calls themselves are quite light, just pulling and pushing data from and to a MySQL db.
Over 1,000 users could be logged in concurrently (up to 10,000 in the near future). It doesn't seem to be an issue, until 100+ or so users log in concurrently, or near-concurrently.
I then get the server clogged with tens of stuck PHP-CGI processes.
Can anyone share advice on an adequate FastCGI config for this scenario? I am - to say the least - confused by
Maxrequests vs Maxinstances Activity timeout vs Idle timeout Queue length Rapid fails per minute vs Request timeout Using tcp vs using Namedpipe
Thanks