Glossary

Some common terms have specific meanings in the context of Fivetran documentation.

Account

When we talk about your account, we mean your Fivetran account. You can manage your account using your Fivetran dashboard. When we need to talk about a source or destination account, we call it out by name (for example, your Jira account or your BigQuery account). See also Fivetran dashboard.

Alert

Alerts appear on your Fivetran dashboard to give you important information about your Fivetran account. There are two types of alerts: errors and warnings. Errors inform you about actions you must take to fix your connectors or transformations. Warnings inform you that something is wrong but is not disrupting your syncs. See also error and warning.

Connector

A Fivetran connector is a data pipeline that moves data from your source to your destination. For example, a Salesforce connector moves your data from Salesforce to your destination. Sometimes you might have multiple connectors of the same type. For example, you might have multiple Google Sheets connectors, each moving data from a different Google Sheet.

Core tables

Fivetran identifies certain tables in the source as mandatory to ensure data accuracy in the connectors. These core tables fetch elementary data from the source that are essential for the connector's efficient functioning. Each Fivetran connector has a different sync strategy that determines the tables that are compulsory and must be included in the sync. The core tables comply with the ERD of the connectors. You cannot deselect or block the core tables from the Schema tab of your connector dashboard.

Cursor

The cursor is the marker that lets us know where the last Fivetran sync left off in your source data. When we start the next sync, we use the cursor to decide where to begin syncing again. Think of the cursor as a metaphorical high-water mark that shows where our sync got to.

The cursor takes different forms depending on the source. For example, in Salesforce, it refers to the last updated timestamp on a particular endpoint, and in PostgreSQL, it refers to an entry in the database's Write-ahead Logs.

Destination

NOTE: A destination was previously known as a data warehouse.

Fivetran connectors replicate your source data to a destination system. Fivetran currently supports the following destination types:

• cloud data warehouses
• databases
• online data platforms
• data lakes

See our documentation for a list of supported destinations. You can have multiple destinations in your Fivetran account. All connectors and transformations are assigned to a particular destination. Each destination is mapped on a 1:1 basis to its group, which is used for user management within the destination. This also means that a destination and its group have the same ID. When you create and manage destinations in your Fivetran dashboard, we automatically manage the associated group. When you create and manage destinations using our REST API, you manage connectors and transformations using our Destination Management REST API resource destination, and users - using our Group Management REST API resource.

Error

NOTE: An error was previously known as a task.

An error is a type of alert in your Fivetran dashboard that tells you about an action you must take to fix your connectors or transformations. We also generate a notification email to let you know about the error. Fivetran creates an error when the problem with your connector or transformation is caused by something that’s on your side. For example, if you have set insufficient permissions in your source and Fivetran can’t sync your data, we generate an error that tells you about the problem and what permissions you must set. See also alert and warning.

Fivetran data models

Fivetran data models are pre-built, open-source SQL statements provided by Fivetran and used to transform your raw data. These models are designed to simplify and accelerate the transformation of raw data into an analytics-ready format by leveraging the source data schemas of Fivetran connectors. You can import Fivetran data models into your dbt project and execute them like any other dbt project. See also Transformation and Quickstart data models.

Fivetran dashboard

Your Fivetran dashboard is the web-based control center for your Fivetran account. Your dashboard provides a comprehensive overview of your account details, including all your connectors and billing information. From your dashboard, you can create and edit connectors, manage your destination, add transformations, add and delete users, review logs, view alerts, and much more. The best way to learn about the dashboard is to explore it yourself. Your view of the dashboard varies depending on your user permissions. Navigate to https://fivetran.com/dashboard/ in any browser and log in to your Fivetran account to access your Fivetran dashboard.

Group

A group maps users to the destination. Each group is mapped on a 1:1 basis to its destination. This also means that a destination and its group have the same ID. When you create and manage destinations in your Fivetran dashboard, we automatically manage the associated group. When you create and manage destinations using our REST API, you manage connectors and transformations using our Destination Management REST API resource destination, and users - using our Group Management REST API resource.

Historical sync

During a historical sync for a connector, Fivetran connects to your source and copies the entire contents of every table that you’ve selected to sync. Historical syncs include all your selected data, including data that is old. How long the historical sync takes depends on the amount of data and the limitations of your source. For example, some sources only allow a limited number of API calls.

A connector's first-ever historical sync is called an initial sync. See also sync.

Historical sync time frame

Some connectors allow selecting how many months' worth of data is synced during their historical sync. This is called the historical sync time frame. The time frame varies by connector type. Some connectors support modifying the time frame even after the connector has been set up, while other connectors do not.

History mode

History mode is a sync mode that Fivetran uses to track history. In this mode, we record every version of each record in the source table to the corresponding table in your destination.

Depending on the connector type, Fivetran uses different approaches to retain historical data:

• For connectors where Fivetran defines the schema, we track history for a predefined connector-specific set of tables. We select those tables based on the analytical value of their historical data. For these connectors, you cannot change which tables track history.
• For connectors where Fivetran just replicates the schema, we give you the option to select, which tables track history. For these connectors, you can switch any supported table to history mode to track its history.

For more information on history mode, see our documentation.

Incremental sync

NOTE: Incremental sync is also known as incremental update.

Incremental syncs update only new or modified data. Fivetran connectors sync most tables using incremental updates. We use a variety of mechanisms to capture the changes in the source data, depending on how the source provides change data. During incremental syncs, Fivetran maintains an internal set of progress cursors, which let us track the exact point where our last successful sync left off. Incremental syncs are efficient because they update only the changed data, instead of re-importing whole tables. See also sync.

Initial sync

An initial sync is a connector's first-ever historical sync. When you first create a connector and sync the historical data in your source, we don’t charge you for the historical data. However, the incremental data may count towards paid MAR, depending on the scenario.

NOTE: If you choose to block or hash columns before running the initial sync for your new connector, we query your data source and cache your data while we fetch the full schema. We write to the destination only the data you selected as tables and columns from the fully fetched schema and only when you approved the selection.

Lite connector

A Fivetran Lite connector is a type of connector built for specific use cases with an accelerated development cycle.

Normalize

When Fivetran normalizes data, we organize the data into tables and columns in a way that reduces data redundancy and stores it logically. Normalization divides larger tables into smaller tables and links them using relationships, according to specific rules.

Quickstart data models

Quickstart data models is a no-code implementation of Fivetran data models. It allows you to use the data models without needing to manually write or edit code. See also Transformation.

Re-sync

A full re-sync completely overwrites the data in your destination with new data from your source. A table re-sync lets you overwrite the data in a specific table so that you can fix data integrity issues in selected tables without re-syncing the entire connector. Normally, Fivetran uses incremental updates to sync data from your source to your destination, so we only sync data that has changed. However, sometimes discrepancies occur between the data in your destination and your source. Then you need to overwrite existing data in your destination to make it consistent with the source, and a re-sync lets you do that. Read more about our re-sync feature. See also sync and incremental sync.

Re-import

If we cannot sync a table incrementally (for example, due to a missing corresponding incremental endpoint), we must sync it in full during every sync. We call this operation a re-import. Re-import tables incur more MAR than incremental tables.

When we know a re-import table is slow or problematic, we flag that table as Not Recommended on your connector's Schema tab. Some connectors with many re-import tables, like NetSuite SuiteAnalytics, have a special logic to only sync re-import tables once a day or once a week.

Rollback sync

A rollback sync is a sync that automatically runs once a day. It captures the changes that happen outside the incremental sync time frame, including metrics that may have been revised after your last sync. Revision is common practice for many applications that generate reports. See also sync.

The rollback window is how many days' worth of data we request in each sync. For example, with a rollback window of 30 days, the connector syncs all data from the past 30 days. Some connectors have default rollback windows; others allow you to customize your rollback window within a certain range (for example, between 2 and 90 days). To ensure data integrity, we align our rollback windows with the time frame in which each source application revises its data.

Schema

A database schema defines how the data is organized in a database. It contains the different tables, their fields, and the relationship between tables. When you create a new Fivetran connector, you choose the name that the schema will have and Fivetran creates it in your destination. For most sources, each connector results in one schema in your destination. Database sources are the exception because a single database connector can replicate multiple schemas. Read more about how database connectors handle multiple source schemas.

Source

A source is a specific database, application, file storage service, event tracking service, or function from which you wish to sync your data. Some examples of sources are MongoDB, Salesforce, or Google Sheets. Sometimes we will pull data in different ways for a single source, such as Adobe Analytics with our unique Adobe Analytics and Adobe Analytics Data Feed connectors. This isn't very common, but in this case we also refer to each way of pulling data as a unique source.

Sync

During a sync, Fivetran replicates data from your source and loads it into your destination. Fivetran does not retain sync data except in specific circumstances. Learn how Fivetran syncs work in our Sync Overview documentation. See also historical sync, incremental sync, initial sync, re-sync, and rollback sync.

Transformation

Transformations are SQL scripts that are executed on your data based on specific events or conditions. Transformations map incoming data into a specific shape that is easier or faster to use in the next part of your data pipeline. Fivetran uses the term "transformation" to refer to two different types of reshaping:

• Pre-load transformations: Fivetran performs some minor transformations on your data before we load it into your destination.

• Post-load transformations: Fivetran offers a feature called Transformations, which supports custom transformations in the destination after your data is loaded.

Fivetran offers Transformations for dbt Core*. Read our Transformations documentation for details. See also Fivetran data models

Transformation Scheduling

Transformation scheduling defines when and in what order transformations are executed. Scheduling allows transformations to run at designated times, in response to specific events, or at regular intervals, ensuring that the data is always up-to-date and ready for analysis.

Warning

Warnings are a type of alert in your Fivetran dashboard that tell you about a problem with your connector that you may need to fix but that is not disrupting your data syncs. For example, a warning might tell you that we were unable to sync specific tables or columns because you are still using a column that has been deprecated by your application's API. See also alert and error.

* dbt Core is a trademark of dbt Labs, Inc. All rights therein are reserved to dbt Labs, Inc. Fivetran Transformations is not a product or service of or endorsed by dbt Labs, Inc.