ALM Toolkit String error

[ronap]

Hello all,

I have very recently installed the ALM toolkit for Power BI and still figuring out how to use it! I made a few changes to columns within my data model to update on the service (changed some column names and added a calculated column). The toolkit update worked the first time I used it (I changed some column names as a test and this updated the model in the service perfectly fine). 

However upon using it again I keep receiving the following error:

The JSON DDL request failed with the following error: <ccon>Cannot convert value '4/28/2024 9:42:26 AM' to required type 'String'. Check path 'model.tables[15].partitions[0].source.refreshBookmark', line 6596, position 64.</ccon>

I've been unable to locate the error, and I don't currently have any date columns formatted to strings. I have been sifting through the tables and have been unable to find the value specified (some of them have > 50 columns so it's been a little difficult).

Is anyone able to point me in the right direction? I've assumed line 6596, position 64 refers to row 6596 and column 64 though this hasn't been much help, I think the data in the preview in PQ is sorted differently. Is anybody able to explain what exactly model.tables[15] will be referring to? Or explain the different pats of this error message to help me pinpoint the underlying issue - I'm a little new to this so any help will be very appreciated!

Thanks you !

[hackcrr · Accepted solution]

Hi, @ronap 

The Instruct JSON DDL (Data Definition Language) command failed because it encountered values that could not be converted to strings. Specifically:
model.tables[15]: refers to the 16th table in the model (indexed from zero). partitions[0]: refers to the first partition of the table. source.refreshBookmark: refers to the attribute in the source definition of the partition that is the refreshBookmark. Line 6596, position 64: indicates the exact location in the JSON file where the error was detected.

Troubleshooting Steps
In Power BI Desktop, go to the Model view. Count the tables and identify the 16th table. Note the table name.
Check Partitions: If you are using incremental refresh, your table may have multiple partitions. In Power BI Desktop, you can check partitions in the Model view under Table Properties.
2. Locating errors in JSON
Export JSON Definition: exports the JSON definition of the model. This can be done using a tool such as the table editor or by exporting the model schema in Power BI Desktop.
Search for refreshBookmark: Open the exported JSON in a text editor.
Search to refreshBookmark to find a specific path and the values associated with it.
3. Checking the refreshBookmark value
The error indicates that the value model.tables[15].partitions[0].source.refreshBookmark at, '4/28/2024 9:42:26 AM' which is a DateTime, cannot be directly converted to a string.
4. Correct the value
If refreshBookmark should be a string, make sure the value is formatted correctly as a string. It may be necessary to update or correct the value in the refreshBookmark model definition or ensure that ALM Toolkit handles the value correctly.
5. Validating Data Types
Ensure that all date columns in the model are correctly formatted as dates and are not inadvertently stored or formatted as strings.

The following is a general example of JSON correction:

{
  "model": {
    "tables": [
      {
        "name": "YourTable",
        "partitions": [
          {
            "source": {
              "refreshBookmark": "2024-04-28T09:42:26Z" // ISO 8601 format
            }
          }
        ]
      }
    ]
  }
}

Best Regards,

hackcrr

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.