We have a repo server running on a Windows box serving some static files that are supposed to be grabbed by our worker nodes using Python-urllib. The problem lies in the fact that when we make a request to a file that we know exists using Python we receive a 404 error, however, if I access the file directly from my browser, using the same link that Python-urllib is using, it works perfectly fine, and the file is downloaded. Example access log:
xxx.xxx.xxx.xxx - - [24/Jul/2013:17:15:53 -0400] "POST /x/dl/Bukkit_Beta.jar.conf HTTP/1.1" 405 172 "-" "Python-urllib/2.7"
I did notice something weird in error.log:
2013/08/04 22:29:11 [error] 6456#2696: *807 CreateFile() "C:\Users\Administrator\Desktop\Nginx/html/MCProHosting/dl/405" failed (2: The system cannot find the file specified), client: 198.15.64.226, server: localhost, request: "GET /MCProHosting/dl/405 HTTP/1.1", host: "repo.mcprohosting.com"
It does appear that from these logs a 405 is being generated, rather than a 404, but Python shows a 404 page. Please note that this is a fresh download of nginx the same directory structure works under Apache if we transfer over the htdocs/html folder.
No configuration changes have been made to Nginx