Overview

MoEngage allows you to import users and events through tables in your Databricks databases.

Types of Imports

MoEngage supports the following types of imports from your Databricks data warehouse:

Registered Users: Users who are already registered on MoEngage.
Anonymous Users: Users who are not yet registered on MoEngage.
Events (standard and user-defined): MoEngage can import standard events like campaign interaction events and your user-defined events.

Prepare the Data

MoEngage does not require a specific table schema for imports; all columns in the table can be skipped or mapped individually on the MoEngage dashboard. However, certain considerations must be addressed before configuring the imports.

User Imports
Event Imports

When users configure periodic User Imports on MoEngage, MoEngage syncs data that has been changed since the last synchronization by referencing the updated_at timestamp column. You can assign any name to this column as long as it accurately reflects the time of the data modification. If you’re assigning this column by a different name in your table, you can configure this mapping separately on the MoEngage dashboard.

To import standard MoEngage events, ensure that the event names in your table align with MoEngage standard event names.

Required Access Permissions

MoEngage requires READ access to your database so that we can fetch data into MoEngage. You can grant the following permissions to an existing database user or create a new dedicated database user for MoEngage:

Query 1 (Required)

GRANT SELECT ON SCHEMA catalog_name.schema_nameTO[email protected]; The query above grants the SELECT permission on all tables within the schema schema\_name in the catalog catalog\_name to the user with the email [email protected]. Granting the SELECT permission enables MoEngage to perform the following actions within the schema:

Read data from all tables within the schema.
Execute SELECT queries on any table.
View the contents of tables without the permission to modify the underlying data.

Query 2 (Required)

GRANT USE SCHEMA ON SCHEMA catalog_name.schema_nameTO[email protected]; The query above grants the USE SCHEMA permission on all tables within the schema schema\_name in the catalog catalog\_name to the user with the email [email protected]. The USE SCHEMA permission grants a user the ability to:

Access and view the schema’s metadata.
Set the schema as their active working context (for example, using USE SCHEMA).
View the schema in schema listings.
Utilize the schema name when referring to fully-qualified object names.

When you use both Query 1 and Query 2, the permissions function as follows:

USE SCHEMA: This permission is required to access and reference the specified schema.
SELECT: This permission allows data to be read from tables within that schema.
Without the USE SCHEMA permission, a user cannot access the schema, even if they have SELECT permissions on its tables. Query 2 explicitly grants the USE SCHEMA permission to enable schema access.

In all the queries provided, ensure you replace the following placeholder values with your specific details:

<catalog_name>: The name of your catalog.
<schema_name>: The name of your database/schema.
<[email protected]>: The email ID of the user who created the token.

Import Datetype Attributes

Importing Datetype attributes requires additional steps. For more information, refer here.

Set Up Imports from Databricks

Prerequisites

Ensure you have an existing Databricks connection setup in the MoEngage App Marketplace with relevant permissions.
If your security policies require you to whitelist our IPs, you can refer to here.
Support for Object Data Type needs to be enabled for your account - Optional.

To set up Databricks Imports, perform the following steps:

On the sidebar menu in MoEngage, hover over the Data menu item . The Data menu appears.
Click Data imports.
On the Data imports page, click the Data warehouses tab.
Click + Import in the upper-right corner and select Users or Events to create a new import.
Click the Databricks tile.
Click Continue.

Step 1: Select Your Databricks Connection and Table Source

Import Name

Enter a name for this import to identify it on the Imports Dashboard easily. Based on the type of import selected, your next steps might vary:

User Imports
Events Imports

You can now select whether to import Registered users or Anonymous users. You can also choose to import both together:

For Event imports, you can select which event it is that you want to import:

You can either select an existing event or create a new one. If you have multiple events stored in a single table, MoEngage determines matching rows to import based on the Event Name column. The value of the Event name column is determined by the event you pick. Some of these events might have different Display name and Event name. You can always view the event information on the Data Management

page. For example, the event App/Site Opened must have the value of the Event name column as MOE\_APP\_OPENED:

To create a new event, perform the following steps:

Click the + Create event at the end of the Select event list. The Create new event dialog box is displayed.
In the Event name box, type a unique name for your event. By default, your Display Name will be the same as the Event Name.

You can go to the Data Management page to view or edit this event and your other MoEngage events.

The newly created events do not appear on the Data Management page until the initial import is successful.

Import Source

In this first step, Source and format, you must specify MoEngage, which Databricks connection to use, and the table from which to import. To get started, perform the following steps:

In the Databricks connection list, select a connection to use for this Import. If you have not already created a Databricks connection, click + Add connection at the end of the Databricks connection list, and you will be redirected to the App Marketplace to set it up. You can learn more about connecting your Databricks warehouse to MoEngage here.
After you have selected your Databricks connection, the Schema/Dataset and Table/View lists are displayed.
In the Schema/Dataset, select the schema/dataset. Note: Ensure that MoEngage has been granted the necessary permissions detailed in the Prerequisites if your schemas are loading incorrectly.
In the Table/View list, select table/view to import data from.

Event Imports

In addition to the above steps, MoEngage provides additional support for tables containing multiple events. If your table contains multiple events, you must first Preview the table and then select the Table contains multiple events check box.

MoEngage uses the values of the Event name column to filter out rows that need to be imported. It imports only those rows that match the selected event name. You can designate an existing column in your table as the event name column. After selecting this column and previewing the data again, filtered rows are displayed for your review before proceeding with the import:

After you preview your table, you will move to the second step, Import configuration and action.

Step 2: Map Your Columns to MoEngage Attributes

In this step, you need to map the columns of your table to the attributes present in MoEngage. All your columns are shown one below the other:

Column name: This specifies the column name to be mapped. Below the column name, MoEngage also displays a sample value (picked from the first row of the fetched table in the previous step) for your reference.
Map attribute: This specifies the MoEngage attribute you want to map the table column to. You can also choose to create a new attribute. Some attributes support ingestion from multiple data types, so you need to pick the column’s data type as well. For the datetime columns, you must pick the format. For more information, refer here.
Action: You can optionally choose to skip the column. The skipped column will not be imported.

Depending on the type of import, there are a few mandatory mappings required:

User Imports

Registered Users
Anonymous Users
All Users

Mapping	Description
User ID	In your table, include a column with a unique user identifier, which is essential for identifying user accounts within your system.
Updated at	MoEngage uses this column to determine which rows have been added/updated since the last sync. You must ensure that this timestamp (date+time) is in UTC Timezone. The column type for this should be `TIMESTAMP`. For the complete list of supported datetime formats, refer to this section.

Mapping	Description
Anonymous ID	You need to mark a column from your table (email, mobile number, etc.) as an identifier of these users.
Updated at	MoEngage uses this column to determine which rows have been added/updated since the last sync. You must ensure that this timestamp (date+time) is in UTC Timezone. The column type for this should be `TIMESTAMP`. For the complete list of supported datetime formats, refer to this section.

Mapping	Description
User ID	Your table must have a column containing a unique user identifier (that identifies a user’s account in your system). Any user with an empty user ID automatically gets imported as an anonymous user.
Anonymous ID	You must mark a column from your table (email, mobile number, etc.) as an identifier of these users.
Updated at	MoEngage will use this column to determine which rows have been added/updated since the last sync. You need to ensure that this timestamp (date+time) is in UTC Timezone. The column type for this should be `TIMESTAMP`. For the complete list of supported datetime formats, refer to this section.

Event Imports

Mapping	Description
User ID	This column matches MoEngage user IDs to your events.
Event time	Map the column that contains the timestamp (date + time) of when the event occurred. MoEngage uses this column to sync new events. Ensure that this timestamp (date + time) is in the UTC time zone. The column type must be `TIMESTAMP`. The Event Time of the imported event will be converted to the timezone chosen in your MoEngage dashboard settings. For a complete list of supported datetime formats, refer to this .
Updated on	MoEngage uses this column to determine which rows were added or updated since the last sync. Use this column to import backdated events or manage delays in upstream pipelines. Ensure that this timestamp (date and time) uses the UTC time zone. If you do not have this column, select Use same as Event time. The column type must be `TIMESTAMP`. For the complete list of supported datetime formats, refer to this section.

After a mandatory mapping is marked, it will reflect against the column name in the mapping table, and you can no longer mark the column as skippable. Just like new events, you can also create a new user attribute. To do so:

Click + Create attribute available in the Select attribute list. The Create new attribute dialog box is displayed.
In the Attribute name box, type a name for your attribute.
In the Data type list, select a data type. You can edit this and existing attributes from the Data Management page.
The newly created user attributes will not appear on the Data Management page until the initial import is successful.

Manifest Files

Optionally, you can choose to auto-map these columns by uploading a Manifest file. To upload a manifest file:

Click the Upload mapping file in the upper-right of the mapping table.
On the Upload mapping dialog box, upload your manifest file.
Click Done.

Your mappings are auto-configured accordingly. Any columns with non-MoEngage attributes are left blank, and you can either manually map the column or create a new attribute for it. Make sure your manifest file follows the expected conventions. Any additional columns in your Manifest File that are not in your table will be ignored. Also, if the mapping for an existing table column is not present in the manifest file, MoEngage will keep the mapping blank so that you can manually configure it.

If a column in the manifest file is mapped to a non-existent MoEngage attribute, the mapping will be blank, and you will need to manually create a new attribute from the UI and then map it.

Support for Object Data Type

The Object data type is supported in Databricks as well.

Store Compatible JSON Data in Databricks

To store JSON data inside Databricks, you must change the data type of the column to VARIANT type. For more information, refer here. The JSON stored inside Databricks should be a valid JSON; otherwise, the values will not be written as JSON. Here is an example JSON column:

{ "Designation": "SSE", "Palace": "Banglore", "age": 30, "name": "Shasha" }

Import JSON Data via Databricks

The JSON data can be imported into Databricks by associating existing attributes in the MoEngage platform that have been designated as Object type with columns in Databricks.

You can also create new Object attributes by clicking Create new attribute in the Select attribute list under the Map attribute column.

MoEngage does not support mapping with nested attributes. Only top-level attributes are available to map.

Save Users as a Segment

When importing users, you can include them in a custom segment in MoEngage. The imported users are consistently added to this segment with each sync, and no users will be removed. To save imported users as a custom segment, perform the following steps:

Turn the Save as a custom segment toggle on to save your imported users in a custom segment and send tailored campaigns to the same.
In the Segment name box, type a name for your segment.
In the Column having user ID list, select the Identifier column in your table.

Import Behaviour

In the case of User Imports, you can also choose to update existing users only. This is helpful when you want to bulk update users’ attributes in MoEngage without creating any new users. To enable this, select the Update existing users only check box under Import Behaviour:

Send Import Notifications

You can choose to be notified about the status of your imports via email. To do so:

Select the Send import status to check box.
In the Select email id list, select the email ID. You can select up to 10 emails to send the status emails to.

The import status email contains info about the following events:

An import was created
An import was successful
An import failed

After completing all mappings, click Next.

Step 3: Select the Import Frequency

In this step, you must define when to sync with your tables. We support the following types of imports:

One-Time: You can run the import as soon as possible or at a later date and time (scheduled). All existing rows that match the import criteria are imported.
Periodic: You can run your imports hourly, daily, weekly, or monthly, or with intervals and advanced configurations.

Optionally, you can specify whether the import should end after a specified set of occurrences or at a particular date. Click Done when ready.

Import schedule and detail times shown on the Data Imports dashboard use your app’s configured timezone. If no app timezone is set, they display in UTC by default. Set the app timezone from Settings to see these times in your preferred timezone. This display timezone is separate from the mapped Updated at and Event time columns, which must always be in UTC.

Upon initial import, all matching rows from your table are imported. Subsequent imports only include changed rows.

Duplicate Imports

A duplicate import is considered when the:

Users/ Events import types are the same.
Event name/ Registered/ Anonymous/ All users import subtypes are the same.
Event name/ Registered/Anonymous/All users import types with the same Databricks connection.
Event name/ Registered/Anonymous/All users import types with the Schema/Dataset and Table/View are the same.

Import Failure Policy

In the event of a connection failure, a recurring Databricks import will retry up to 10 times. If all retries fail, the import is marked FAILED. The following table outlines the behavior and required action when a recurring Databricks import fails due to connection issues (for example, credentials, network, or warehouse unavailability).

Policy	Description	Action
Automatic Retry	MoEngage automatically attempts to restore the connection and retries the import up to 10 times in the event of a connection failure.	None (MoEngage handles automatically)
Import Failure	If all 10 retries fail, MoEngage marks the import as FAILED and stops all further automatic attempts and future imports. After it has failed, you must create a new import by either duplicating the failed import or creating a fresh one.	Mandatory Investigation and Restart

Restart a Failed Import

Any import that has FAILED requires your manual intervention:

You must investigate and resolve the underlying issue on your Databricks side (for example, update credentials, check permissions, or ensure warehouse availability).
After resolving the issue, you must manually duplicate the import from the MoEngage UI to restart it.

The first successful execution of any new recurring import is always a historical import (it fetches all the data from the configured table/view).
Duplicating a failed import initiates a new historical import. If the original import previously succeeded, this duplication may result in redundant data. To control data flow effectively, MoEngage recommends using views.

FAQ

Does MoEngage support Databricks Unity Catalog connections, and how does the setup differ?

Answer: Yes, MoEngage’s Databricks import is built upon the Databricks Unity Catalog. Therefore, the setup process is the same as a standard Databricks connection.

Are there specific version requirements for Databricks or Databricks SQL Warehouses for MoEngage integration?

Answer: No, there are no specific version requirements for compatibility.

What should the Databricks column data type be for JSON data to import successfully as an Object Data Type into MoEngage?

Answer: Databricks requires the column type to be VARIANT during table creation.

How does MoEngage handle Databricks TIMESTAMP columns containing local timezones instead of UTC during import?

Answer: MoEngage interprets stored time as UTC, as Databricks lacks a local time concept. Databricks advises storing timestamps in UTC.

​Overview

​Types of Imports

​Prepare the Data

​Required Access Permissions

​Query 1 (Required)

​Query 2 (Required)

​Import Datetype Attributes

​Set Up Imports from Databricks

​Step 1: Select Your Databricks Connection and Table Source

​Import Name

​Import Source

​Event Imports

​Step 2: Map Your Columns to MoEngage Attributes

​Manifest Files

​Support for Object Data Type

​Store Compatible JSON Data in Databricks

​Import JSON Data via Databricks

​Save Users as a Segment

​Import Behaviour

​Send Import Notifications

​Step 3: Select the Import Frequency

​Duplicate Imports

​Import Failure Policy

​Restart a Failed Import

​FAQ

Overview

Types of Imports

Prepare the Data

Required Access Permissions

Query 1 (Required)

Query 2 (Required)

Import Datetype Attributes

Set Up Imports from Databricks

Step 1: Select Your Databricks Connection and Table Source

Import Name

Import Source

Event Imports

Step 2: Map Your Columns to MoEngage Attributes

Manifest Files

Support for Object Data Type

Store Compatible JSON Data in Databricks

Import JSON Data via Databricks

Save Users as a Segment

Import Behaviour

Send Import Notifications

Step 3: Select the Import Frequency

Duplicate Imports

Import Failure Policy

Restart a Failed Import

FAQ