v1 Warehouse Sync API tutorial

Use this tutorial to configure your first Warehouse Sync pipeline with the mParticle Postman collection, and activate the data in a downstream tool. Postman is an easy and friendly environment for developers and non-developers alike to use APIs.

This tutorial is not a complete guide to all of the Warehouse Sync features and APIs. For more reference documentation, see the Warehouse Sync API Reference.

Prerequisites

• Install the latest version of the Postman desktop application. You can download Postman from https://www.postman.com/downloads/.

• Fork the mParticle Warehouse Sync Postman Collection to your workspace:

  

  A copy of the Warehouse Sync environment is included. You can download it again here.

Step 1. mParticle setup

Create Platform API credentials

You need credentials to use the Platform API to create Warehouse Sync Pipelines.

To create a Platform API credential:

1. After signing in to the mParticle app as a user with the Admin role, click the user icon in the bottom left corner.

   
2. Click Settings.
3. Click the API Credentials tab.
4. Click the green Add Credential button at the top right of the screen.

5. Give the credential a name, and ensure the Platform checkbox is checked and the Admin dropdown is selected. Click the green Save button to save the new credential.

   
6. Copy the Client ID and Client Secret values, as you’ll need these in a later step.
7. Click Done.

Step 2. Data warehouse setup

Work with your warehouse administrator or IT team to ensure your warehouse is reachable and accessible by the mParticle app.

1. Whitelist the mParticle IP address range so that your warehouse will be able to accept inbound calls from mParticle.
2. Ask your database administrator to execute the following steps in your warehouse to create a new Role that the mParticle service will use to connect to and access your database. Make sure to select the correct tab for your warehouse (Snowflake, Google BigQuery, or Amazon Redshift).

USE ROLE ACCOUNTADMIN;

// Create a unique role for mParticle 
CREATE ROLE IF NOT EXISTS {{role_name}};

GRANT USAGE ON WAREHOUSE {{compute_wh}} TO ROLE {{role_name}};
GRANT USAGE ON DATABASE {{database}} TO ROLE {{role_name}};
GRANT USAGE ON SCHEMA {{database}}.{{schema}} TO ROLE {{role_name}};

// Grant SELECT privilege on any tables/views mP needs to access
GRANT SELECT ON TABLE {{database}}.{{schema}}.{{table}} TO ROLE {{role_name}};

// Recommend creating a unique user for mParticle
CREATE OR REPLACE USER {{user_name}} PASSWORD = "{{unique_secure_password}}";
GRANT ROLE {{role_name}} TO USER {{user_name}};

// Note: Don't change the name of the storage integration: plug in your mP POD, Org ID, 
// and Acct ID into the name, and be sure to leave the other characters as they are 
// ("mp_", underscores, and "_s3").
//
// Use the value for your mParticle instance for {pod_mp_aws_account_id}:
// US1 = 338661164609
// US2 = 386705975570
// AU1 = 526464060896
// EU1 = 583371261087
CREATE OR REPLACE STORAGE INTEGRATION mp_{{pod}}_{{org_id}}_{{acct_id}}_s3
          WITH TYPE = EXTERNAL_STAGE
          STORAGE_PROVIDER = S3
          ENABLED = TRUE
          STORAGE_AWS_ROLE_ARN = "arn:aws:iam::{{pod_mp_aws_account_id}}:role/ingest-pipeline-data-external-{{org_id}}-{{acct_id}}"
          STORAGE_AWS_OBJECT_ACL = "bucket-owner-full-control"
          STORAGE_ALLOWED_LOCATIONS = ("s3://{{pod}}-ingest-pipeline-data/{{org_id}}/{{acct_id}}");


GRANT USAGE ON INTEGRATION mp_{{pod}}_{{org_id}}_{{acct_id}}_s3 TO ROLE {{role_name}};

DESCRIBE INTEGRATION mp_{{pod}}_{{org_id}}_{{acct_id}}_s3;

-- Create a unique user for mParticle
CREATE USER {{user_name}} WITH PASSWORD '{{unique_secure_password}}'

-- Grant schema usage permissions to the new user
GRANT USAGE ON SCHEMA {{schema_name}} TO {{user_name}}

-- Grant SELECT privilege on any tables/views mP needs to access to the new user
GRANT SELECT ON TABLE {{schema_name}}.{{table_name}} TO {{user_name}}

{
    "Statement": [
        {
            "Action": "sts:AssumeRole",
            "Effect": "Allow",
            "Resource": "arn:aws:iam::{{pod_mp_aws_account_id}}:role/ingest-pipeline-data-external-{{org_id}}-{{acct_id}}",
            "Sid": ""
        }
    ],
    "Version": "2012-10-17"
}

Step 3. Postman setup

Once you have installed Postman, configure the collection environment.

Update Postman environment settings

1. Ensure you have forked mParticle Warehouse Sync Postman Collection in the Prerequisites section. In Postman, click the Environments tab on the left hand navigation.

2. Once successfully forked, you’ll see it in the list of Environment configurations. You can rename it to something more meaningful by right-clicking on the … next to the name and choosing the Rename option.

   

3. Replace the placeholder value “REPLACE_ME” for several environment variables with values corresponding to your environment. Ensure you update the values in the CURRENT VALUE column.

   • Replace PLATFORM_API_CLIENT_ID and PLATFORM_API_CLIENT_SECRET with the Platform API credentials from the mParticle Setup Section.

   • Replace WORKSPACE_ID, ACCOUNT_ID, and ORG_ID with the corresponding values.

     • To find your Org ID, log into the mParticle app. View the page source. For example, in Google Chrome, go to View > Developer > View Page Source. In the resulting source for the page, look for “orgId”:xxx. This number is your Org ID.
     • Follow a similar process to find your Account ID (“accountId”:yyy) and Workspace ID (“workspaceId”:zzz).
   • Replace POD with the POD your mParticle account is deployed on. Look at the URL in your browser where you are signed into mParticle. The POD is one of the following values: US1, US2, EU1, AU1.
   • Replace PASSWORD with the password of the warehouse user account you defined in the Data Warehouse Setup section.

4. Enter the corresponding data warehouse username and password you copied down in “Step 2. Data Warehouse Setup.” according to the data warehouse you are using.

   • For Snowflake, replace SNOWFLAKE_AWS_IAM_USER_ARN and SNOWFLAKE_AWS_EXTERNAL_ID with the values you copied from step 2.
   • For BigQuery, replace SERVICE_ACCOUNT_ID with the service account ID you used in BigQuery, and replace SERVICE_ACCOUNT_KEY with the key from the generated JSON file in step 2.
   • For Redshift, replace REDSHIFT_AWS_IAM_ROLE_ARN with the ARN of the role created in step 2.
5. After updating all the values, click COMMAND-S (or CTRL-S) to save your changes.

Update the Postman collection

1. Ensure you have forked the mParticle Warehouse Sync Postman Collection in the Prerequisites section. In Postman, click the Collections tab on the left hand navigation.
2. Once successfully forked, you’ll see the collection in the list of available collections.

3. Click the Warehouse Sync collection to expand it. Then click the Variables tab.

   

4. Replace placeholder values (the ones indicated with “REPLACE_ME”) for several variables with the values corresponding to your environment. Ensure you update the values in the CURRENT VALUE column.

   • Replace INGEST_PIPELINE_SLUG and INGEST_PIPELINE_NAME with values for what you’d like to call the pipeline you’ll be creating.

   • Replace SQL_QUERY with the database query mParticle should use to retrieve the data from your warehouse. SQL is a powerful language and you can use advanced expressions to filter, aggregate, join, etc. your data. Work with your database administrator if you need help crafting the right SQL query:

     • Your query can contain a timestamp column that mParticle will use to keep track of which rows need to be loaded.
     • Your query should contain one or more user identity columns that mParticle will use to perform identity resolution to ensure that data ends up on the correct user profile.

     • As part of the SQL query, you must specify how columns in the query will map to attributes on a user’s profile. You do this by using column aliasing in SQL. For example, in the following query, the column cid in Snowflake is being mapped to the mParticle attribute customer_id.

       

       If you don’t provide an alias, mParticle will use the name of the column in your database. If an attribute of this name does not already exist on the user’s profile, mParticle will create a new attribute with this name.

   • Before using the query in mParticle, test the query outside of mParticle to ensure it is returning the data you think it will.
   • To learn more checkout the Warehouse Sync SQL reference
5. After updating all the values, click COMMAND-S (or CTRL-S) to save your changes.

Step 4: Create your first Warehouse Sync pipeline

Creating a Warehouse Sync pipeline takes four steps:

1. Create the partner feed.
2. Create the connection.
3. Create the data model.
4. Create the pipeline.
5. Monitor the pipeline.

Create the partner feed

The first step is to create a feed. mParticle uses this information for rules, and connections in your workspace. By submitting a request to the Feeds API. You will need to provide a value for module_name that corresponds with the data warehouse you are configuring with Warehouse Sync. Valid values for module_name are Redshift, BigQuery, or Snowflake. Note that these values are case sensitive. For example, if you use snowflake instead of Snowflake, you will encounter errors later in your configuration.

1. In Postman, ensure the environment drop-down is pointed to the Environment configuration you recently imported.
2. Expand the Warehouse Sync Collection and open the Feeds folder.
3. Click Create Warehouse Sync Feed.

4. Click the Body tab to see the information you will pass to the API in order to create the feed.

   
   • module_name can be one of Snowflake, BigQuery, or Redshift

   The values in {{Sample Values}} will be taken from the variables you updated in previous steps.

5. Once you are confident all values are correct, click the blue Send button to issue the API request that creates the feed. If everything is correct, mParticle returns a success message with details about the feed you just created. If the request was not successful, mParticle returns an error message with additional information about what the issue might be.

   

Create the connection

The next step is to create a connection. mParticle uses this information in order to establish a connection with your data warehouse.

1. In Postman, ensure the environment drop-down is pointed to the Environment configuration you recently imported.

   
2. Expand the Warehouse Sync Collection and open the 1) Connections folder.
3. Click POST Create Connection.

4. Click the Body tab to see the information you will pass to the API in order to create the connection.

   

   The values in {{Sample Values}} is taken from the environment variables you updated in earlier steps. Make sure these values match the values for your organization’s data warehouse. You may need to work with your database administrator to ensure you have the correct values.

5. Once you are confident all values are correct, click the blue Send button to issue the API call to create the Connection configuration. If everything is correct, mParticle will return a success message with details about the configuration you just created. If it was not successful, you will get an error message with additional information pointing to what the issue might be.

   

Create the data model

The second step in the pipeline creation process is creating a data model. mParticle uses this information in order to determine what data should be extracted from your database and how it will map to mParticle’s data model.

1. In Postman, expand Warehouse Sync Collection and open the 2) Data Models folder.
2. Click the Create Data Model request.

3. Click the Body tab to see the information you will pass to mParticle in order to create the data model.

   

4. The values in {{Sample Values}} will be taken from the variables you updated in previous steps. If your SQL Query provided in an earlier step includes a timestamp field, update the values for load_timestamp_field_type, load_timestamp_field_name, and load_timestamp_field_time_zone to match. They should be blank if your SQL Query does not contain a timestamp.

   • load_timestamp_field_type is the warehouse data type of your timestamp column.

     • The timestamp values for Snowflake are: timestamp_ltz, timestamp_ntz, timestamp_tz, datetime, date, timestamp_unixtime_ms, timestamp_unixtime_s.
     • The timestamp values for Google BigQuery are datetime, date, timestamp, timestamp_unixtime_ms, timestamp_unixtime_s.
     • The timestamp values for AWS Redshift are date, timestamp, timestamptz, timestamp_unixtime_ms, timestamp_unixtime_s.
   • load_timestamp_field_name is the timestamp column name in your SQL query.
   • load_timestamp_field_time_zone indicates whether the field has an associated time zone. Specify a valid IANA timezone value here (ex: America/New_York). — List of IANA values.

5. Once you are confident all values are correct, click the blue Send button to issue the API request that creates the data model. If everything is correct, mParticle returns a success message with details about the data model you just created. If the request was not successful, mParticle returns an error message with additional information about what the issue might be.

   

Create the pipeline

The final step in the process is creating the pipeline itself. You will pass in the Connection and Data Model configurations previously created and provide a few additional details.

1. In Postman, expand Warehouse Sync Collection and open the 3) Pipelines folder.
2. Click the request called Create Pipeline.

3. Click the Body tab to see the information you will pass to mParticle in order to create the Pipeline configuration.

   

   The values in {{Sample Values}} will be taken from the variables you updated in previous steps. No values are required to be updated in this step. You can optionally update the environment variable. It is currently set to target your mParticle development environment, but you can change it to target your production environment.

4. You can associate an optional data plan with your data warehouse pipeline to help improve the quality of, and grant you more control over, the data imported to mParticle from your warehouse. In the request body of your call to the Data Warehouse Sync API, include the body parameters "plan_id" and plan_version and set these to the values of the data plan and version you want to use. You must use a data plan version that is active and that exists within the workspace you are creating the pipeline.

5. Once you are confident all values are correct, click the blue Send button to issue the API request that creates the Pipeline configuration. If everything is correct, mParticle returns a success message with details about the configuration you just created. If it was not successful, mParticle returns an error message with additional information about what the issue might be. If you kept the default value for the scheduled_interval (set to “once”), the pipeline kicks off immediately after creation and only runs once. You can set it hourly, daily, or weekly.

   

Monitor the pipeline

Once a pipeline has been created, you can monitor its status using the additional requests provided in the Postman collection.

1. In Postman, expand the Warehouse Sync Collection and open the 3) Pipelines folder.
2. Click the Get Pipeline Status request.
3. Click the blue Send button to submit the request.

4. mParticle sends a detailed message with the pipeline’s current status. Note that after creating a pipeline, there is an approximate one-minute delay until the pipeline is created in the mParticle backend, so submitting a Get Pipeline Status call results in a Not Founderror. Try again in a minute if you experience this.

   

5. While a Pipeline is ingesting data, monitor it in mParticle, just like you do with other inputs. From mParticle, go to Data Master > Live Stream to inspect the incoming data from Snowflake.

   

6. Once the data is ingested, you will see the data points appear on the User’s Profile. Go to Activity > User Activity and look up a sample profile. If the attributes do not show up the way you thought they would, validate the mapping/aliasing you provided in the SQL Query provided earlier.

   

Step 5. Activate the data

Now that the data has been loaded into mParticle, it’s time to activate it.

For this section of the tutorial, we will create an audience that uses the loaded data and sends it to a downstream integration.

Create an audience

Create an audience that uses one of the attributes we ingested as a qualifying criteria for it:

1. In the mParticle app, go to Audiences > Standard.
2. Click the green New Standard Audience in the upper right corner.
3. Give the audience a name.
4. Select a date range, or choose “All available data.”
5. Select the Warehouse Sync feed you created in mParticle Setup.

6. Add an audience criteria that leverages one of the data points you ingested from your warehouse. In the example below, we only want to consider users who have a propensity-to-buy score that’s greater than 0.7.

   
7. Add any other criteria you want to be considered. Click Save As Draft when you are done.
8. Click Calculate to run the audience calculation.

Connect the audience to an output

In the final step, we will send users who qualified for this audience to Iterable for further targeting. If your organization doesn’t use Iterable, pick a different integration that your organization uses.

After the audience has been fully calculated, connect it to an output:

1. If you aren’t still there, go to Audiences > Standard.

2. In the row for the audience you just created, click the Connect button in the Actions column.

   

3. Click Connect Output.

   
4. Click the Add Output green button.
5. Click the Iterable tile in the integrations directory, or pick another integration your organization uses if you don’t have Iterable.

6. Select the Audience checkbox and click Configure.

   

7. Enter a configuration name, your Iterable API key, and the user ID that is used to identify users, and then click Save & Open in Connections.

   

8. Provide the ID of the list in Iterable that the user data should be loaded into, and click Add Connection.

   

9. Click the Send button to send the audience to Iterable.

   

10. Data starts flowing to Iterable.

    

Now you can open the audience in Iterable and take the next step.