locked
Request Fails On Verizon Cellular Connection, Same Request Succeeds On WiFi and Other Cellular Connections RRS feed

  • General discussion

  • My company is experiencing some issues with a particular request on Verizon. For reasons of privacy I am not posting the request here, however it is getting a stream from a CDN provider, with headers setting the UserAgent to a custom value as requested by the client (who are filtering at the server end for this UserAgent).

    On Verizon cellular data connection in the Northern part of the USA only, the request fails with

    HTTP/1.1 400 Bad Request
    Server: Footprint 4.8/FPMCP
    Mime-Version: 1.0
    Date: Thu, 01 Jan 1970 00:00:00 GMT
    Content-Type: text/html
    Content-Length: 840
    Expires: Wed, 30 Jul 2014 19:40:21 GMT
    Proxy-Connection: close

    The request succeeds without issue on WiFi and all other cellular providers (such as AT&T), and when tested succeeds with Verizon cellular data connection in the South-West of the USA.

    If we remove the UserAgent from the call to be the default NativeHost then all calls/requests fail with the exact same error on all networks. 

    Other requests made by the WP8 application which require this UserAgent set to a custom value all succeed on Verizon in the Northern part of the US, so we have pretty much eliminated Verizon stripping out the UserAgent as it passes through their network.

    The same call succeeds without issue (with a different platform specific UserAgent value) on other platforms.

    Our client has stated that they do not operate any geo-specific servers so the location shouldn't be a factor (but in reality it does seem to be), and also that they have the UserAgent filtering correctly deployed on their servers.

    Has anyone else experienced similar issues / oddities with requests on WP8 apps on Verizon. If so how were they resolved?


    Wednesday, July 30, 2014 11:03 PM