Modeling status reports

The Modeling status reports file allows you to track the modeling progress for each scenario that is included in your Intelligent Recommendations account's feature set. The file is located in the Storage Account logs folder.

You can see a copy of the Modeling Status Reports for your storage account by following these steps:  

  1. From Azure portal, go to your Storage account, select Containers and locate your ircontainer (or whatever container you use for hosting the data contract). 
  2. Select the ir_logs folder.
  3. You now see a folder named StatusReport. You notice that each modeling cycle status is reported in a different file within this folder, and kept separate from your error-logging folder. The file name includes file's last update time and the modeling cycle ID (also known as environment ID).

Note

If you are not seeing error logs in your ir_logs folder, check that the connection string for your model is configured correctly.

Status reports  

The status reports file contains the following parameters: and potential status types:

Status Report Parameter Description
executionID Modeling cycle ID.
startTimeUtc Represents the start time of the modeling cycle.
reportLastUpdateTimeUtc Represents last update time of the Status reports file, and is updated hourly.
scenariosStatus List of all algorithms represented in your model and the associated model status (success, in progress, failed).

The status reports file contains the following potential status types:

Status Description
Success This status means the modeling process of the scenario is finished and you can view full recommendation results with the API for this scenario.  
InProgress This status means the modeling process of the scenario is still running. In some cases recommendation results might be partially ready at this stage. In order to get full recommendation results, you should wait for the modeling process of the scenario to finish with ‘Success’ status.
Failed This status means that there has been an error and the scenario has failed in the modeling stage. To investigate further, start by consulting the ir_logging folder to see if there are errors with the data contract. After the changes are made, manually trigger the model and try again. A failed algorithm won't impact the results of other algorithms.

The following table maps each algorithm to its impacted API scenario(s):

Modeling Status Algorithm Related API Scenario(s)
TextualSimilarity Represents the model that powers textual recommendations, created using your Similar API with the algoType="Textual"
VisualSimilarity Represents the model that powers Visual Similarity recommendations, which include:
AlgoType="visual"
BringtheLook
CompleteSimilarItems
CompleteSimilarStyles
Picks Represents the model that powers the "Picks for you" recommendations (Picks API).
NextBest Represents the model that powers the NextBestAction API (formerly CART).
PeopleAlso Default Similar API response
BrowseLists Represents the status of your trending chart recommendations (New, Trending, Popular) and ReRankedAlgoLists.
ExternalLists Represents the status of your External Lists and ReRankedExternalLists.

Example status report

In this example, your BrowseLists and ExternalLists have finished cooking, but your NextBestAction API and PeopleAlso API requests are still being modeled.

{
"executionId":your_execution_id,
"startTimeUtc":"2022-09-03T21:25:24Z",
"reportLastUpdateTimeUtc":"2022-09-08T09:08:59.9493937Z",
"scenariosStatus": {
    "NextBest":{"status":"InProgress"},
    "PeopleAlso":{"status":"InProgress"},
    "ExternalLists":{"status":"Success"},
    "BrowseLists":{"status":"Success"}
  }
}

See Also

Quick start guide: Make an authenticated API call
Deployment overview
Configure Azure Data Lake Storage reports
Use data contracts to share data