Open APIs

Expand all | Collapse all

[TMF653] Bug/Feature that ServiceTest.validFor is writeable

  • 1.  [TMF653] Bug/Feature that ServiceTest.validFor is writeable

    TM Forum Member
    Posted Mar 05, 2021 01:54

    Hi Folks,

    I noticed that within the current TMF653-ServiceTest-v4.0.0.swagger.json the ServiceTest.validFor attribute is writeable thru ServiceTest_Create.validFor and ServiceTest_Update.validFor.
    The description states:

    The validity time for the test results

    Does that make sense? I would assume an API-Consumer should only read this attribute after the Service Test has been completed. Is this simply a minor error or is there a backstory to it?

    Cheers,

    Jan



    ------------------------------
    Jan Lemmermann
    OSS Lead Architect
    EWE TEL GmbH
    ------------------------------


  • 2.  RE: [TMF653] Bug/Feature that ServiceTest.validFor is writeable

    TM Forum Member
    Posted Mar 05, 2021 08:29
    Service Test is led by @Johanne Mayer and @Ernie Bayha - they might be able to provide additional insights.
    As a general rule, however, I would advocate having as few restrictions as possible in an API model, the obvious ones such as that href and id are not patchable.
    The ability or not to update a specific field would be restricted using role-based access control.
    Hope it helps.

    ------------------------------
    Jonathan Goldberg
    Amdocs Management Limited
    Any opinions and statements made by me on this forum are purely personal, and do not necessarily reflect the position of the TM Forum or my employer.
    ------------------------------



  • 3.  RE: [TMF653] Bug/Feature that ServiceTest.validFor is writeable

    TM Forum Member
    Posted Mar 09, 2021 01:55

    Thanks @Jonathan Goldberg makes sense if this is a general pattern.

    Maybe @Johanne Mayer or @Ernie Bayha can confirm that this is intentional.

    Regards,

    Jan​

    ​​

    ------------------------------
    Jan Lemmermann
    OSS Lead Architect
    EWE TEL GmbH
    ------------------------------



  • 4.  RE: [TMF653] Bug/Feature that ServiceTest.validFor is writeable

    TM Forum Member
    Posted Mar 09, 2021 06:41
    Hi, I am working on some enhancements on TMF653 so I will look into it.

    Regards.... Johanne

    ------------------------------
    Johanne Mayer
    MayerConsult Inc
    ------------------------------



  • 5.  RE: [TMF653] Bug/Feature that ServiceTest.validFor is writeable

    TM Forum Member
    Posted Mar 09, 2021 19:04
    Hi Jan,  I read your comment again to make certain I was not missing anything.  The ServiceTest.validFor attribute is provided for the create and update operations to provide the time that these test results will be considered valid.  Similar to a quote that is valid for a period of time some tests may be valid also for a period of time if the producer of the test knows that nothing will change in the next 24 hours for example.  The producer would provide the length of time the test would be valid for in the execution.  It could be used at the access level to test the speed or whatever other measure you want to test.

    If I missed understood your question just let me know.

    Regards... Johanne

    ------------------------------
    Johanne Mayer
    MayerConsult Inc
    ------------------------------



  • 6.  RE: [TMF653] Bug/Feature that ServiceTest.validFor is writeable

    TM Forum Member
    Posted Mar 10, 2021 02:54

    Hi Johanne,

    thanks a lot for the feedback. I think for us the question came up because of our use case.

    System-A (API Consumer) wants to run a service test in System-B (TMF653 API Provider) and executes a CREATE of a /serviceTest.
    System-B executes the test and reports back in the service test with "validFor" how long the test results are valid. In our case System-B has this knowledge, because it executes the test.

    In our case the API Consumer can't specify the value, because it doesn't know the details about where the test result data comes from and how long they are valid.
    But I can also understand if the permission level is kept rather "open". You can still decide in the concrete implementation whether an API consumer can write the field.

    I hope this explains a little bit the background of my question :-)

    Regards,

    Jan



    ------------------------------
    Jan Lemmermann
    OSS Lead Architect
    EWE TEL GmbH
    ------------------------------