dbt Core

An overview of the dbt Core integration with Secoda

dbt Core Metadata Extracted

Getting started with dbt Core

dbt is a secondary integration that adds additional metadata on to your data warehouse or relational database tables. Before connecting dbt make sure to connect a data warehouse or relational database first. These include Snowflake, BigQuery, Postgres, Redshift, etc.

There are several options to connect dbt core with Secoda:

1. (Recommended) Connect an AWS, GCP, or Azure storage bucket/container

2. Upload a manifest.json and run_results.json through the UI

3. Upload a manifest.json and run_results.json through the Secoda API

Option 1 – Storage bucket (container)

This option is recommended to ensure that Secoda always has the latest manifest.json and run_results.json files from dbt Core. Secoda will only sync these files from the bucket.

1a. Connect an AWS S3 bucket

You can connect to the AWS S3 bucket using an AWS IAM user, or AWS Roles.

AWS IAM User

1. Create a new AWS IAM user and ensure that Access Key - Programatic access is checked. Once you create the user save the Access Key ID and Secret Access Key that are generated for the user.

2. Attach the following policy to the user. Make sure to change <your-bucket-name>.

{
    "Statement": [
        {
            "Action": [
                "s3:PutObject",
                "s3:PutObjectAcl",
                "s3:ListBucket",
                "s3:GetObject",
                "s3:GetObjectAcl",
                "s3:DeleteObject"
            ],
            "Effect": "Allow",
            "Resource": [
                "arn:aws:s3:::<your-bucket-name>",
                "arn:aws:s3:::<your-bucket-name>/*"
            ]
        }
    ],
    "Version": "2012-10-17"
}

1. Connect your S3 bucket to Secoda

   • Navigate to https://app.secoda.co/integrations/new and click dbt Core

   • Choose the Access Key tab and add the credentials from AWS (Region, Bucket Name, Access Key ID, Secret Access Key)

   • Test the Connection - if successful, you'll be prompted to run your initial sync

AWS Roles

1. Create a new AWS IAM role. In the Select type of trusted entity page, click Another AWS account and add the following account ID: 482836992928.

2. Click on Require External ID, and copy the randomly generated value from Secoda, in the dbt Core connection page.

3. Attach the following policy to the role. Make sure to change <your-bucket-name>.

{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Effect": "Allow",
            "Action": [
                "s3:GetObject",
                "s3:ListBucket"
            ],
            "Resource": [
                "arn:aws:s3:::<your-bucket-name>",
                "arn:aws:s3:::<your-bucket-name/*"
            ]
        }
    ]
}

1. Once the role is created, you'll receive an Amazon Resource Name (ARN) for the role.

2. Connect your S3 bucket to Secoda

   • Navigate to https://app.secoda.co/integrations/new and click dbt Core

   • Choose the Role tab and add the credentials from AWS (Role ARN, Region, Bucket Name)

   • Test the Connection - if successful you'll be prompted to run your initial sync

1b. Connect a GCS S3-compatible bucket

1. Login to GCP cloud console.

2. Create a service account.

3. Grant access to the service account from the Bucket page as “Storage Object Viewer”.

4. Turn on interoperability on the bucket. Generate HMAC keys for a service account with read access to the bucket. Both located here:

1. Setup CORS. GCP requires this be done over CLI. Like the following:

gsutil cors set cors.json gs://bucket-name

cors.json

[
  {
    "origin": ["*"],
    "method": ["GET", "POST", "PUT", "DELETE", "OPTIONS"],
    "responseHeader": ["Content-Type"],
    "maxAgeSeconds": 3600
  }
]

1. Save the HMAC keys to be used in the connection form.

   • Access Key Id

   • Secret

   • Region bucket region for GCP

   • S3 Endpoint must be added and set to https://storage.googleapis.com

2. Connect your S3 bucket to Secoda

   • Navigate to https://app.secoda.co/integrations/new and click dbt Core

   • Choose the Access Key tab and add the HMAC keys saved above to the relevant fields.

   • Test the Connection - if successful you'll be prompted to run your initial sync

1c. Connect a Azure Blob Storage container

1. Go to portal.azure.com and then click Storage accounts.

2. Copy the name of the desired storage account. Enter that in the integration form.

3. Click on your storage account and under Security + networking select Access keys.

4. Copy the Connection string and add to your integration form.

5. Test the connection.

Option 2 – Upload a single manifest.json

The dbt manifest file contains complete information about how tables are transformed and how they are connected in terms of data lineage. It details the model to table relationships, providing a complete and accurate lineage view.

This is a one time sync with your manifest.json file. You can upload the file following these steps:

1. Navigate to https://app.secoda.co/integrations/new and click dbt Core

2. Choose the File Upload tab and select your manifest.json and run_results.json files using the file select

3. Test the Connection - if successful you'll be prompted to run your initial sync

Option 3 – Using the API

The API provides an endpoint to upload your manifest.json and run_results.json file. This is convenient if you run dbt with Airflow because you can upload the manifest.json at the end of a dbt run. Follow these instructions to upload your manifest.json via the API:

1. Create a blank dbt core integration by going to https://app.secoda.co/integrations/new and selecting the "dbt Core" integration and then click "Test Connection". And run the initial extraction. This extraction will fail, but that's intended.

2. Return to https://app.secoda.co/integrations and click on the dbt Core integration that was just created. Save the ID which is contained in the URL.

3. Use the endpoints below to upload your files.

Endpoints ->

3a. Two separate calls (One for Manifest, One for Run Results)

Manifest.json: https://api.secoda.co/integration/dbt/manifest/

Run_results.json: https://api.secoda.co/integration/dbt/run_result/

Method -> POST

NOTE -> This will automatically trigger an extraction to run on the integration you created

Sample Request for Manifest file (Python) ->

import requests

headers = {
    "Authorization": "Bearer <Your Key>"
}
response = requests.post(
	"<https://api.secoda.co/integration/dbt/manifest/>",
	files={"manifest_file": open("manifest.json", "rb")},
	data={"integration": "Your Integration ID"},
	headers=headers
)
print(response.json())

Sample Request for Run Results file (Python) ->

import requests

headers = {
    "Authorization": "Bearer <Your Key>"
}
response = requests.post(
	"<https://api.secoda.co/integration/dbt/run_result/>",
	files={"run_result_file": open("run_results.json", "rb")},
	data={"integration": "Your Integration ID"},
	headers=headers
)
print(response.json())

3b. One call to upload both Manifest and Run Results

1. Get your dbt Core Integration ID

• Get the integration ID from the integration page URL

  • For example, if the url is https://app.secoda.co/integrations/f7d68db5-9dbc-4880-b6cd-ec363c1f7d6b/syncs, the integration id would be f7d68db5-9dbc-4880-b6cd-ec363c1f7d6b

• Or get the integration ID programmatically via a GET request to the /integration/integrations/ endpoint and parse the list for your dbt Core integration

2. Upload manifest.json and run_results.json

Endpoint (inserting the integration_id from Step 1): https://api.secoda.co/integration/dbt/{integration_id}/upload_artifacts/

Method -> POST

Expected Response -> 200

Sample Request for uploading your files (Python, note the TODOs) ->

import requests

# TODO: replace YOUR_INTEGRATION_ID with your integration_id
integration_id = "YOUR_INTEGRATION_ID"

url = f"https://api.secoda.co/integration/dbt/{integration_id}/upload_artifacts/"

# TODO: replace YOUR_PATH with your path to run_results and manifest
files={
  "run_results": open("YOUR_PATH/run_results.json", "rb"), 
  "manifest": open("YOUR_PATH/manifest.json", "rb")
}

# TODO: replace YOUR_API_KEY with your Secoda API key
headers = {
  "Authorization": "Bearer YOUR_API_KEY"
}

response = requests.request("POST", url, headers=headers, files=files)

print(response)

3. Trigger an Integration Sync

Endpoint (inserting the integration_id from Step 1): https://api.secoda.co/integration/dbt/{integration_id}/trigger/

Method -> POST

Expected Response -> 200

Sample Request for triggering a sync (Python, note the TODOs) ->

import requests

# TODO: replace YOUR_INTEGRATION_ID with your integration_id
integration_id = "YOUR_INTEGRATION_ID"

url = f"https://api.secoda.co/integration/dbt/{integration_id}/trigger/"

# TODO: replace YOUR_API_KEY with your Secoda API key
headers = {
  "Authorization": "Bearer YOUR_API_KEY"
}

response = requests.request("POST", url, headers=headers)

print(response)

4. Now that a sync has been triggered, you can now monitor your sync in the UI!