Release Version 2.46

Last updated on Jul 22, 2026

In this Release

✨ New and Changed Features 🔧 Fixes and Improvements 📄 Documentation Updates

New and Changed Features

Sources

Support for Automatically Restarting Historical Load after OpLog Expiry in MongoDB (Added in Release 2.45.2)

New

Introduced the option to automatically restart the historical load if the Oplog or Change Stream expires in your Pipeline. This option is available during Source configuration and when editing an existing Pipeline. When enabled, Hevo restarts the historical load for all active objects, preventing Pipeline failures and eliminating the need for manual intervention.

MongoDB →

Support for Company-to-Company Association Object in HubSpot

New

Added support for ingesting data from the Company to Company association object. This object captures relationships between companies in your HubSpot account, such as parent–subsidiary associations.

To ingest data for this association object in new and existing Pipelines, contact Hevo Support to enable it.

HubSpot →

Support for the Latest LinkedIn Marketing API

Enhanced

Effective Mar 09, 2026, the LinkedIn Ads integration is upgraded to use Version 202601 to ingest data. This upgrade introduces support for retrieving data from the Posts API, enabling ingestion of posts associated with LinkedIn Ads sponsored accounts. The Posts object is now supported, and the Video Ads object has been deprecated and replaced by it.

The upgrade process will be seamless, with no downtime for your Pipelines. To ingest the historical data for the Posts object in an existing Pipeline, you must restart the historical load for the object.

LinkedIn Ads →

Support for Version 2026-01-15 of the Klaviyo v2 API

Enhanced

Effective Release 2.46, Hevo has updated the Klaviyo v2 integration to support API version 2026-01-15. This update expands subscription consent handling and updates the structure of certain Campaign and Message responses, including additional subscription channels and consent types for Profiles, sending strategy and tracking fields on the Campaign API, and support for mobile push campaigns.

This upgrade was applied seamlessly and did not require downtime or configuration changes. This change applies to all new and existing Pipelines.

Klaviyo v2 →

User Experience

Hevo SaaS Offering on Snowflake Marketplace (Added in Release 2.45.1)

New

Introduced the Hevo SaaS offering via Snowflake Marketplace, enabling customers to subscribe with customized pricing, contract terms, or payment schedules. Existing Snowflake Marketplace Capacity Drawdown (MCD) credits can be used toward the purchase.

Customers can contact Hevo Support to request a private offer.

Fixes and Improvements

Sources

✓

Handling Events Loading Issue in LinkedIn Ads (Fixed in Release 2.45.1)

Fixed an issue where Events were not loaded for the Ad Analytics object in some Pipelines. This issue was introduced in Release 2.43.3, when Hevo added support for the audience_penetration and average_dwell_time fields, but did not include dateRange while querying them. Since dateRange is required to derive the date value that forms part of the object’s composite primary key, the Events failed validation and were not loaded. With this fix, Hevo now includes dateRange when querying for these fields, ensuring all Events are loaded as expected. This fix applies to all new and existing Pipelines. To ingest the missing data in existing Pipelines, restart the historical load for the Ad Analytics object.

✓

Handling Pipeline Failures for Sharded Clusters in OpLog Mode in MongoDB (Fixed in Release 2.45.1)

Fixed an issue where Pipelines could be created with sharded MongoDB clusters using OpLog mode, even though this configuration is not supported. This issue occurred because each shard in a sharded cluster maintains its own OpLog, making it difficult for Hevo to read and manage them, causing such Pipelines to fail during ingestion. With this fix, Hevo no longer allows you to create Pipelines that use sharded clusters with OpLog mode. To create a Pipeline with a sharded cluster, use Change Stream mode instead.