timeout and re-tries

leor.nussel
leor.nussel Newcomer
in DSS

Hi


  • After how many seconds do you declare a timeout and what is the response code for it?
  • What is your best practice regarding re-tries? After how long and how many re-tries do you recommend?

thanks

Tagged:

Best Answer

  • Christiaan Meihsl
    Christiaan Meihsl LSEG
    Answer ✓

    @leor.nussel,

    I'm not sure what exactly you mean by a timeout ?

    Most DSS API calls are asynchronous. This means that if the requested data is not available after 30 seconds they will return an HTTP status 202, and a URL you can poll to find out when the extraction is complete. The DataScope Select REST API conforms to the OASIS OData standards for async. You can find more information in the Key Mechanisms page on Async.

    Notes:

    • Our tutorials and code samples all handle the async mechanism.
    • If you are using the .Net SDK all this is handled transparently in the background by the SDK.
    • Polling to find out when an extraction is complete: choose a reasonable polling interval, to avoid placing a useless load on the network and servers. Anything less than 30 seconds should be avoided. I'd recommend 1 minute or more.

    You might find the Best Practices and Limits document useful in this context.

Answers

  • Gurpreet
    Gurpreet admin

    @leor.nussel There is no need for a timeout; all server failures will be communicated back to your application in the response message; and OS / language libraries will communicate any local network failures. For very long running extraction queries, you can opt for asynchronous delivery of data.

    Please see the sample implementation in code examples for DSS in C#, Python and Java language in the downloads tab.

Categories