locked
Help with an error code from Request/Response RRS feed

  • Question

  • I am trying to invoke my model from the api endpoint that was built.  I am getting this message:

    Internal Server Error;System has encountered an error, when reporting this issue, please mention request ID c35a2884-c640-4d9f-8949-0a4a96adba62

    Here is the calling method:

    staticasyncTaskInvokeRequestResponseService()

            {

               

    using(varclient = newHttpClient())

                {

                   

    ScoreDatascoreData = newScoreData()

                    {

                        FeatureVector =

    newDictionary<string, string>()

                        {

                            {

    "Project_id_withsr", "p08-8568a"},

                            {

    "Proj_Type", "Project"},

                            {

    "AppID", "A100"},

                            {

    "Release_Month", "October"},

                            {

    "Release_Year", "2013"},

                            {

    "Capability_Pattern", "Hybrid"},

                            {

    "Domain_Proj", "Member"},

                            {

    "Domain_App", "Medical Management"},

                            {

    "Compliance_ind", "N"},

                            {

    "SLA_Class", "3"},

                            {

    "DR_Class", "1"},

                            {

    "Pre_Imp_Find_Crit", "3"},

                            {

    "Pre_Imp_Find_Maj", "0"},

                            {

    "Pre_Imp_Find_Min", "0"},

                            {

    "Pre_Imp_Find_Cos", "0"},

                            {

    "Pre_Imp_Find_Tot", "0"},

                            {

    "Pre_Imp_Find_Ind", "0"},

                            {

    "Pre_Imp_Find_CritMaj", "3"},

                            {

    "Pre_Imp_Def_Crit", "0"},

                            {

    "Pre_Imp_Def_Maj", "0"},

                            {

    "Pre_Imp_Def_Min", "0"},

                            {

    "Pre_Imp_Def_Cos", "0"},

                            {

    "Pre_Imp_Def_Tot", "0"},

                            {

    "Pre_Imp_Def_CritMaj", "0"},

                            {

    "Pre_Imp_Find_Ratio", "0"},

                            {

    "Post_Imp_Def_Tot", "0"},

                            {

    "Post_Imp_Def_Ind", "0"},

                            {

    "Post_Imp_Def_CritMaj", "0"},

                            {

    "Post_Imp_Def_CritMaj_Ind", "0"},

                            {

    "SysTest_Scripts", "10"},

                            {

    "Perf_Test_Ind", "N"},

                        },

                        GlobalParameters =

    newDictionary<string, string>()

                        {

                        }

                    };

                   

    ScoreRequestscoreRequest = newScoreRequest()

                    {

                        Id =

    "score00001",

    Instance = scoreData

                    };

                   

    conststringapiKey = ""

                    client.DefaultRequestHeaders.Authorization =

    newAuthenticationHeaderValue("Bearer", apiKey);

                    client.BaseAddress =

    newUri("https://ussouthcentral.services.azureml.net/workspaces/19a2e623b6a944a3a7f07c74b31c3b6d/services/5d324f8570bc4d559eef5f0e36c69231/score");

                   

    HttpResponseMessageresponse = awaitclient.PostAsJsonAsync("", scoreRequest);

                   

    if(response.IsSuccessStatusCode)

                    {

                       

    stringresult = awaitresponse.Content.ReadAsStringAsync();

                       

    Console.WriteLine("Result: {0}", result);

                    }

                   

    else


                    {

                       

    Console.WriteLine("Failed with status code: {0}", response.StatusCode);

                    }

                }

    Thanks in advance

    Tuesday, September 23, 2014 9:58 AM

Answers

  • Hi Jamie,

    I found the root cause for your original error (c35a2884-c640-4d9f-8949-0a4a96adba62): Train dataset must have more than one observation

    Are you trying to publish a web service that include Train Model? Generally you will find that the workflow in Azure ML is to train in one experiment, right click -> Save As Trained Model on the output of Train Model, and use that saved model in another experiment that only has a Score model.

    I can walk you through this if that wasn't clear

    Regards,

    AK

    • Marked as answer by Jamie.DixonMVP Wednesday, September 24, 2014 12:17 PM
    Tuesday, September 23, 2014 2:01 PM

All replies

  • I went ahead and removed all columns except 1 string column and the dependent variable column (int).  It still throws an error. 

    Service call failed. Error 500 (Internal Server Error;System has encountered an error, when reporting this issue, please mention request ID 61101740-07dd-4ea0-a2b7-19e3c332989d) when requesting /webservicegroups/7ca9881db6e545d5a0c6c7044dc1335c/webservices/aa48bcb6715e4792b836b24e8af08aea/score

    Tuesday, September 23, 2014 10:32 AM
  • Hi Jamie,

    I found the root cause for your original error (c35a2884-c640-4d9f-8949-0a4a96adba62): Train dataset must have more than one observation

    Are you trying to publish a web service that include Train Model? Generally you will find that the workflow in Azure ML is to train in one experiment, right click -> Save As Trained Model on the output of Train Model, and use that saved model in another experiment that only has a Score model.

    I can walk you through this if that wasn't clear

    Regards,

    AK

    • Marked as answer by Jamie.DixonMVP Wednesday, September 24, 2014 12:17 PM
    Tuesday, September 23, 2014 2:01 PM