none
HTTP Error 503 Retry-After header value format RRS feed

  • Question

  • Hi there!

    According to the Exchange ActiveSync protocol specifications (https://msdn.microsoft.com/en-us/library/gg650846(v=exchg.80).aspx):

    The server returns an HTTP error 503 when more users than are allowed by the server's request queue limit have sent requests to a single server or when the actions of the client have triggered throttling.

    If a Retry-After header ([RFC2616]) is present in the response, the client SHOULD<6> retry the request after waiting the number of seconds indicated by the Retry-After header. Any such value represents an estimate of when the server is expected to be able to process the request.

    However, our email client's logs indicate that when connecting to Office 365 using Exchange ActiveSync protocol, the value corresponding to Retry-After header is equal to a non-integer value (e.g. 2398.1874306). This contradicts with the RFC2616 specifications that require:

    The value of this field can be either an HTTP-date or an integer number of seconds (in decimal) after the time of the response.

    Can you clarify this? Is it an error in the Exchange ActiveSync specifications or something wrong with the Office365 server implementation of the specifications?

    Here's the example headers that our client received from the Office 365 server:

    [Cache-Control: private, Content-Length: 27, Content-Type: text/html, Retry-After: 2398.1874306, Server: Microsoft-IIS/10.0, request-id: 2be310d9-44f5-408d-a847-9d34dd9bbec4, X-CalculatedFETarget: CY1PR1201CU001.internal.outlook.com, X-BackEndHttpStatus: 503, X-FEProxyInfo: CY1PR1201CA0008.NAMPRD12.PROD.OUTLOOK.COM, X-CalculatedBETarget: CY1PR04MB1945.namprd04.prod.outlook.com, X-BackEndHttpStatus: 503, MS-Server-ActiveSync: 15.20, X-MS-ASThrottle: SyncCommands, X-MS-BackOffDuration: H/777, X-AspNet-Version: 4.0.30319, X-DiagInfo: CY1PR04MB1945, X-BEServer: CY1PR04MB1945, X-FEServer: CY1PR1201CA0008, X-Powered-By: ASP.NET, X-FEServer: LNXP265CA0058, Date: Wed, 31 Jan 2018 14:58:19 GMT]

    Wednesday, January 31, 2018 3:13 PM

All replies

  • Hi NRLN,

    Thanks for posting your question regarding MS-ASHTTP and Exchange Throttling. 

    The MS-AS* ActiveSync specifications do not address Office365 implementations. Only on-premise Exchange scenarios are documented. As such, the Open Specifications Support team does not provide direct support for using the MS-AS* specifications in the Office365 scenario. However a member of the Open Specifications Support team can attempt a comparison to an on-premise installation of Exchange Server and see whether or not the behavior you are seeing in the Office365 scenario is also occurring in the on-premise scenario, and if so we can investigate the issue fully.

    Sincerely,

    Will Gregg | Open Specifications

    Wednesday, January 31, 2018 8:32 PM
    Moderator
  • Hello NRLN:

    I'll help you with this inquiry. Please start logging on o365 for your device as per instructions on this link; reproduce the issue and send logs to dochelp at Microsoft.com for further analysis.

    Thanks.


    Tarun Chopra | Escalation Engineer | Open Specifications Support Team

    Wednesday, January 31, 2018 10:21 PM
  • Hello NRLN: 

    Can you please send logs to dochelp so that we can look into it?

    Thanks

    Tarun.


    Tarun Chopra | Escalation Engineer | Open Specifications Support Team

    Friday, February 2, 2018 5:51 PM
  • Hello NRLN:

     

    We are able to analyze the issue without the requested logs and acknowledge that we are deviating from RFC2616 w.r.t Retry-After format for on-Prem Exchange server as well. We are working on getting this deviation documented in MS-ASHTTP specification to address the issue. The fraction part; 2398.1874306; you see in time is milliseconds component

     

    Thanks a lot for bringing this issue to our attention.

    Regards


    Wednesday, February 7, 2018 5:21 PM