In the RequestDone event of a TSslHttpCli object, a user is reporting that he's getting Error 10053 but StatusCode is 200 and ReasonPhrase is "OK".
This seems to only happen on his "test" server and not his "production" server. The "test" server has a lot less resources and it's being hit hard with a lot of requests but I'm not sure that explains what's happening.
Why would Error be 10053 but StatusCode and ReasonPhrase indicate no error?
UPDATE: Customer adds: "The test and production servers have same configuration files (Apache + nginx), same settings for SSL/TLS (ciphers...). The difference is that the test server uses self-generated certificate, while the production server has certificate issued by Certificate Authority. Most (if not all browsers) do not like self-generated certificate (or bad or obsolete certificates) and issue warning with such URLs. The problem is that only some links (5000+) got 10053 error, not all (300K+ if I remember well) if I refer to the statistics shown during run, so probably the server load would be part of the problem, not the protocol itself."
Thank you!