Glossary of Terms

Last updated on Sep 27, 2024
On This Page
The glossary contains an alphabetical list of words that you may come across in Hevo’s documentation. Click on a letter to jump to the relevant section of the alphabet.

A B C D E G H I J K L M N O P Q R S T U V W X


A

Ad Interaction

The interaction of a user on ad-based Sources, such as a click on an advertisement or a video view.

Ad Squad

A collection of one or more ads where details regarding a Snapchat campaign such as budget, audience, and schedule are defined.

Alteration

A change in the Destination data type to accommodate variations in the Source data type to avoid sidelining of Events for data type mismatches.

Application Programming Interface (API)

An interface that allows two applications to communicate and exchange data. This exchange of data happens through API Endpoints, which act as touchpoints for the APIs to send and receive requests for data. For example, the Google Maps API allows apps and websites to access a variety of Google’s map functions by using the various URLs the API provides for different actions such as:

Display a map — launch Google Maps with no markers or directions:

https://www.google.com/maps/@api=1&map_action=map&parameters

Assertion Consumer Service URL

The URL where the identity provider (IdP or IDP) sends SAML responses. An IdP is a service that stores and manages digital identities. For example, you can use https://auth.hevodata.com/auth/saml/assertion for Hevo to receive SAML responses from your IdP when setting up Hevo as an application in your SAML-enabled IdP.

Auto Mapping

A feature in Hevo that automatically maps new Event Types and their fields to the existing Event Types and their fields in the Destination. Read Auto Mapping Event Types.


B

Back-dated

When the attributes associated with an action are traced to an Event in the past. For example, the conversion on New Year’s Eve was back-dated to the Christmas Day sale.

BinLog

A set of log files in MySQL that record every statement that is executed (adding, deleting, updating records or tables). The log files have a retention period after which they are deleted by the MySQL server. BinLog, short for Binary Log, can be selected as a Pipeline mode for MySQL Sources.

Bookkeeping

The process applied by Hevo Activate on the data fetched from the Warehouse to derive the differences over two successive Activation runs before synchronizing these with the Target object.


C

Candidate Key

Any key or group of keys that uniquely identifies the rows in a table. The candidate key defines which keys can form the primary key for that table. A table can have multiple candidate keys.

For example:

  • Single key: The teacher’s employee ID is a candidate key for the faculty table in the students’ database.

  • Group of keys: The teacher’s name and the subject taught by the teacher together form a candidate key for the subjects table in the students’ database.

Collection

A logical grouping of data within databases like MongoDB or Amazon DocumentDB. It acts as a container that holds multiple documents, providing an organizational structure for storing and managing data.

Combined Index

An index created on multiple columns in a table. For example, consider the following SQL statement:

CREATE INDEX subject_key on SUBJECTS (teacher_name, subject_name);

Here, subject_key is a combined index created on two columns teacher_name and subject_name from the subjects table in the students database.

Composite Index

A synonym for combined index. See Combined Index.

Composite Key

A type of primary key that comprises two or more columns to uniquely identify a record in a database table. For example, a student’s enrollment ID and name in the school’s database.

Connector

A general term for a system that Hevo integrates with another service/software. In Hevo, connectors are Sources, for example, MySQL or Google Ads.

Conversion

The process of an interaction of a user leading to a subsequent purchase or sign-up.


D

Data Type

The form that data exists in. Each field or value has a data type, for example, document, scalar, set, string, or integer.

De-dupe Fields

Fields that can be used to uniquely identify records in Marketo’s standard as well as custom objects. There can be up to three de-dupe fields. In some Activate Targets, a combination of these fields is used as an identifier.

Read Data Synchronization Identifiers.

Debounce

A function used in Hevo search to display suggestions only after a short delay. Instead of showing suggestions each time you type a letter in the search bar, it waits until you have finished or paused typing to show the result, thus improving the user experience.

Deduplication

The act of removing duplicates from the replicated data.

Delivery Semantics

Guarantees the number of times the Events are loaded to the Destination without any data loss. It can be of three types:

  • At-least-once delivery semantics: This ensures that the Events are loaded to the Destination at least once. In case of any failures, Hevo retries to replicate these Events until they are successful.

  • Exactly-once delivery semantics: This ensures that the Events are loaded to the Destination exactly once, without any duplicates, even if there are any failures during the process.

  • At-most-once delivery semantics: This ensures that the Events are loaded to the Destination at most once. In case of any failures, the data is not replicated again, leading to data loss.

Destination

A database, file system, or data warehouse that acts as the endpoint of a data Pipeline, and into which, the data from the Source is finally loaded.

Document

A JSON-like data structure consisting of key-value pairs. A key is a string that represents a field name, and a value can be a string, a number, a boolean, an array, or even nested documents.

Document Database

A type of non-relational database designed to store and query data as JSON-like documents. A document is a record in a Document Database.

Downgrade

When you switch to a subscription plan that has fewer features than your current plan. For example, you downgrade if you switch from a business plan to a starter plan.

Drop (table)

The act of deleting all the data from a table (in the Destination system). This is a permanent action. Hevo does not delete the table itself; users have an option to do so if required.

Dynamic Routing

A network routing method in which routers or routing protocols determine an optimal path for the data packets based on real-time network conditions and changes.


E

Enum Data Type

In MySQL, enum data type is a string object. An enum (enumerated) column can only contain values from a list of pre-defined values specified during table creation.

For example, consider the following table definition:

CREATE TABLE ribbons (
  material VARCHAR(40),
  colors ENUM('red', 'pink', 'blue', 'green', 'black')
);

Here the column, colors can only contain the values red, pink, blue, green, or black.

Event

A fundamental unit of data that represents the creation, update, or deletion of information in the Source and can be replicated to a Destination system. For example, a new document in a MongoDB collection, or updated contact details in the Contacts table in MySQL.

Event Type

Groups of Events created based on the entity they are ingested from at the Source. For example, a table Customers in your Shopify data can be considered an Event Type, and all Events ingested from this table fall under this type.


G

Granularity

The level of detail in a data set. For example, Daily is the lowest time granularity that you can have on Bing Ads API.


H

Historical Data

The data present in the Source application or database before the first ingestion post-Pipeline creation. For some Sources, you can specify the duration for which this existing data must be ingested. However, for other Sources, Hevo may ingest the data either entirely or for a predetermined duration, depending on the Source setup.


I

Immutable Data

The data that cannot be modified after creation. For example, the values assigned to a string in Java.

Incremental Data

The new or modified data in the Source post-Pipeline creation.

Indexes

Special lookup tables that the database search engine can use to speed up data retrieval. An index is like a pointer to the data in a table.

Ingestion

The act of retrieving or fetching data from the Source.

Instance

The data stored in a database at a particular moment in time.

Internet Key Exchange (IKE)

A protocol used in VPN setups to facilitate the creation of secure communication channels between two devices, leveraging authentication keys for enhanced security.

Internet Protocol Security (IPSec)

A set of rules for secure communication over the Internet Protocol (IP) network.

IPSec Tunnel

A secure virtual pathway established between two devices or networks, allowing communication over the public internet via the IPSec protocol.


J

Java Database Connectivity (JDBC)

An SQL-based API created by Sun Microsystems to enable Java applications to use SQL for database access. A JDBC Source is one that supports JDBC connectivity. The JDBC driver for that data source and the URL format is required to set up the connection.

Job

A task in Hevo that replicates data for the objects selected during Pipeline creation, to the Destination. Hevo creates new tasks only when the current running tasks are completed.


K

Keys

A single column or a group of columns that can uniquely identify rows (or tuples) in a table.


L

Latency

The time it takes to load the data into the Destination once it is ingested from the Source. Also called data latency or end-to-end latency.

Logical Replication

The process of replicating data change in a database as per a primary key. Logical replication can be used to perform actions such as, consolidating multiple databases, sending triggers for incremental and individual data change, or sharing a subset of the database between multiple databases.

Lookup Field

A type of field that contains a value retrieved from another object or field. The lookup field can be used to search and associate records between two objects that share a lookup or master-detail relationship. For example, in Salesforce, a user can associate a contact record to an account record using the Account Name lookup field.


M

Master Database

A data repository that contains the server-level settings such as the structure and primary configuration settings of all the databases, and the server-level metadata such as the user authorizations.

Metadata Column

A column that contains information about the data present in a row. For example, __hevo_database_name is a metadata column created by Hevo to indicate the name of the database from which a particular row is ingested.

Read Hevo-generated Metadata to know about the metadata columns created by Hevo.

Metrics

A measure for insights such as clicks, engagement, likes, impressions, followers for your ad campaigns, social media posts or pages.

Mounted State

A database instance in the mounted state is not fully accessible for user connections, and the database is not accessible for normal operations.

Mutable Data

The data that can be modified after creation. Everything except a String type value is mutable by default in Java. For example, the elements in an array can be assigned new values after initialization.


N

Non-nullable Columns

Table columns that do not accept NULL values. This forces a field to always contain a value, which means that you cannot insert a new record or update an existing record, if this column field does not contain a value.


O

OAuth

An open standard for authorization that allows applications or websites limited access to resources hosted by other apps and websites on behalf of the user without sharing the user’s password.

Objects

The entities that Hevo ingests from the Source based on the type of Source. For example, collections from MongoDB or reports from Twitter ads.

Offset

A value in the latest record fetched. Offset is used to identify the starting point for the next set of results to return.

OpLog

A collection present in MongoDB that keeps a record of all the operations that modify the data stored in the database.


P

Parallelism

A process in which when multiple tasks run at the same time. For example, an application can split its tasks up into smaller subtasks that can be processed in parallel, for instance on multiple CPUs at the same time.

Payload

The essential information sent or received with the HTTP methods such as GET or POST, in an API.

For example, in the following JSON response:

Payload Example

The payload is the message, ‘Welcome, world!”.

Pipedrive Pipeline

A framework in Pipedrive that allows you to manage your sales, and keep track of them using multiple stages, from lead arrival, to deal closure. The admin of a Pipedrive account can add these stages, and the team members who you want to allow access to these stages according to their requirements.

Pipeline

A pre-defined framework of user-configured processes in Hevo that move data from one system to another, typically with Transformations that make it easier to analyze the data.

Pipeline Frequency

The frequency at which a Pipeline ingests data from its Source. Read Scheduling a Pipeline.

Poll-based Ingestion

The process by which the data is read from the Source periodically, depending on a set schedule.

Primary Key

A primary key is a non-null candidate key selected to uniquely identify every row in that table. A primary key is a candidate key but the reverse may not be true. For example, the roll number of a student in the students’ marks database uniquely identifies a student.

Primary Server

The server that manages the primary copy of the database, and is responsible for receiving read and write operations directly from clients and executing them against the database. It also sends data to its replica servers.

Primitive Data Type

A data type pre-defined by the programming language. For example, int, long, float, and double in Java.

Private App

An app built by a company in Intercom for its internal use; not listed on the Intercom App Store. For example, ChargeDesk built a private Messenger app for their customer support team to easily share invoices with their customers.

Private Endpoint

An interface that allows private connections between resources without exposing them to the public internet. For example, AWS PrivateLink allows connecting a client in a VPC to an AWS Marketplace service such as MongoDB Atlas.

Proxy Server

An intermediary server that allows multiple clients to route traffic to an external network. A proxy server protects the identity of the connecting client. For example, an organization may have a proxy server set up to route and filter employee traffic to the public Internet.

Public App

An app listed on the Intercom App Store built by a company for use by other companies. For example, Aircall, a cloud-based phone system, is a public app.

Purge

The act of deletion of a BinLog file, an Event, an account, a table, and data in general.

Push-based Ingestion

The process by which Hevo acts as a receiver and the Source holds the responsibility to send/post data to Hevo. This applies to webhook-based Sources.


Q

Query Batch Size

The number of rows or records returned by a SELECT query. The SQL keyword LIMIT is used to define the batch size. For example, consider the following SQL statement:

SELECT id from employees LIMIT 50;

Here, the query batch size is 50. So, the SELECT query returns only 50 records from the employees table.


R

Range Attributes

Values used to specify the minimum and maximum limit within which a field’s value may lie.

Rate Limits

The limits imposed by an API vendor, such as Intercom, HubSpot, on the number of API requests sent to their public APIs. Rate limits only apply on calls to the REST API, applications using OAuth, and Integrations using API keys.

Read-only User

A user who is only allowed to read or access all data from the Source, and not make changes to it.

Read Replica

A copy of the master database that allows read-only requests. They help in disaster recovery and in reducing the traffic or load on the master database by serving the read requests.

Replication

The collective process of ingesting data from a Source application, running Transformations on it, if any is needed, and subsequently loading it to the Destination database or data warehouse.

Reserved Keywords

The special keywords used by SQL relational databases to perform various operations, such as, SELECT and COMMIT. It is not recommended to use these keywords as names for databases, tables, columns, variables, or other objects.

Reverse Proxy Server

A type of proxy server that acts as a gateway between clients, users, and application servers. All incoming client requests are routed via this server to the appropriate server inside the private network. A reverse proxy server protects the identity of the server processing the client request. For example, network administrators can configure the reverse proxy server to whitelist or blacklist specific IP addresses.

RSA key pair

The key pair consisting of a private key and a public key which are used by the RSA algorithm for data encryption. The public key is used for encryption, whereas the private key, which is only available to the user is used for decryption.


S

SaaS (Software as a Service)

A software distribution model where the application is hosted by a company on its servers and is accessed by clients via the internet by paying a subscription fee. For example, Salesforce.

Security Assertion Markup Language (SAML)

A protocol for authenticating web applications. SAML provides a way to authenticate users to third-party web applications such as Gmail for Business, Office 365, and Salesforce.

Scalar

A simple, primitive data type with values like a number or text. For example, int, bool and character in Java.

Schema

The organization of data within an event type which comprises of details such as its name, fields within the event type, and the data types of these fields. For example, the schema of the employees event type may contain fields such as name and ID with datatypes varchar(25) and int respectively.

Schema Mapping

A mapping between fields in the Source schema and the fields of the Destination Schema. It determines the columns in the Destination table into which the incoming data of each Source field must be loaded.

Secure Shell (SSH)

A network protocol that is used to create a secure channel over a network between a client and server application to transfer data.

Segmentation Reporting

A type of reporting applicable to ad-based Sources in which data is retrieved based on metrics segmented by a particular target type. For example, a report segmented by location.

Serverless Computing

A fully-managed platform that provides compute, storage, and network resources on which developers can build and run their applications without having to provision or manage the underlying infrastructure.

Service Account

A Google account associated with a team rather than an individual user. It is user-independent and requires a Key to authenticate the connection. Service accounts are used when workloads are run on virtual machines.

Service Account Key

A public and private RSA key pair which a Google service account uses to securely authenticate the connection.

Set

An interface that implements the mathematical set. A set contains no duplicate elements.

Sidelining

When Events are not loaded in the Destination table due to mismatching of data type with the Source table for the same column.

Signed Data Type

A numeric data type that can hold both negative and positive integers.

Sink

A synonym for Destination. See Destinations.

Snapshot

A data structure that represents the state of the database at a particular point of time.

Source

The application or database from which Hevo ingests the data. For example, GitHub, Facebook Pages, and MySQL.

Subscription Region

The geographical region in which your account is created and its data stored. For example, the subscription region of your Pendo account can be the EU or US region.

Sync

An abbreviated term used to mean synchronization. The terms sync and synchronization are used interchangeably in Hevo’s documentation.


T

Target

Any CRM application to which data may be loaded using Hevo Activate. For example, HubSpot, Salesforce.

Target Object

The table in your CRM applications such as Salesforce and HubSpot with which data is synchronized by an Activation.

Transformation

The process of changing the format and structure of data. Read Transformations.

Truncate

Removes all rows from a table while preserving the table structure, such as its columns, constraints, and indexes. It frees up the storage space used by the table data, making it a faster operation than DELETE, especially for large tables.

For example, consider the following SQL statement:

TRUNCATE TABLE sales_leads;

Here, the data in the sales_leads table is deleted, but the columns and other characteristics are maintained.

Note: This is a non-recoverable operation, which means once the data is truncated, it cannot be recovered.

Tunnel Instance

A virtual machine responsible for managing secured and controlled communication flow between AWS Virtual Private Clouds (VPCs).


U

Unmapped Field

A field that has not been mapped from the Source to the Destination. The Events related to these fields are marked as Failed until you map the fields to the appropriate Destination table fields or skip them. Read Auto Mapping Event Types.

Unsigned Data Type

A numeric data type that can hold only positive integers or zero.

Universally Unique Identifier (UUID)

A data type used to uniquely identify records.

The UUID is written as a sequence of lowercase hexadecimal digits of 128 bits. For example, a0eebc99-9c0b-4ef8-bb6d-6bb9bd380a11.


V

Verified Website

Also known as an HTTPS website that utilizes either Secure Sockets Layer (SSL) or Transport Layer Security (TLS) encryption to secure the information exchanged between the website and its users.

Views

These are virtual tables that does not exist physically in the database but is created by combining data from one or more tables in the Destination. For example, consider the following SQL statement:

CREATE VIEW sales AS
SELECT city
FROM customers
WHERE city = 'New York'

Here, a view called sales is created, which consists of all the customers residing in New York.
Views does not store data in a physical location. It allows you to simplify queries and provide a specific perspective on the data, without actually modifying the underlying data in the tables.

Virtual Private Cloud (VPC) Peering

A private networking connection between two VPCs in cloud computing platforms such as Amazon Web Services (AWS) or Microsoft Azure. It enhances the efficiency of communication among resources in different VPCs.

Virtual Private Cloud (VPC)

A virtual private network (VPN) set up by an enterprise that comprises an on-demand, configurable pool of shared resources allocated within a public cloud environment. A VPN is dedicated to the enterprise creating it and is logically isolated from other virtual networks in the public cloud. For example, Amazon VPC is a VPN dedicated to a user’s AWS account that enables the user to launch AWS resources in that network.

VPC Endpoint Service

A service hosted in an AWS account that can be securely accessed by authorized services in other AWS accounts through AWS PrivateLink.

VPC Interface Endpoint

A type of VPC endpoint from Amazon Web Services (AWS) that acts as a virtual network device. It allows access to a VPC endpoint service and enables private communication between the VPC and supported AWS services.


W

Write-Ahead Logging (WAL log)

A logging mechanism used by PostgreSQL that maintains the transactions taking place in the database. So, even the smallest change in data is written to this log before it is applied to the PostgreSQL database.

Warehouse

A place to store the data accumulated from a wide range of heterogeneous Sources, generally used for data analysis and reporting. For example, Amazon Redshift, Google BigQuery or Snowflake.

Webhook

A methodology that allows a web application to communicate based on an event rather than a request, often with a payload of data.

Workspace

  • A space created on Slack consisting of organized conversations such as channels and DMs, where team members can communicate and work together.

  • A space created in an Intercom account to store users’ data. Users can create different workspaces within an account to organize the data based on their functionality, such as Support, Customer Onboarding, Lead Generation, or Customer Engagement.

  • A space created in Hevo, that allows team members to create and access Pipelines, Models, and Workflows in that workspace. You can add any number of members to your workspace. An organization can create up to five workspaces in Hevo, using the same domain name.


X

XMIN

A system-generated column present in all the PostgreSQL tables.

Tell us what went wrong