Skip to main content

**NEW/IMPORTANT** Ingestion Disruption on Feb 8, 2026 — No Data Loss, Action May Be Required (related to setuptools 82.0.0 update)

Angelo Bresci
Updated

We are writing to inform you of an ingestion service disruption that occurred on Feb 8, 2026, and to provide clear remediation guidance.

What Happened

On Feb 8th, 2026, ingestion failures occurred across most DataHub connectors due to a breaking change in an unpinned third-party dependency. The acryl-datahub package depends on the open-source setuptools library for package management. On Feb 8th, 2026, this library released version 82.0.0 with breaking changes ending support for libraries that DataHub depended on. 

Important: No data loss or data corruption occurred. This issue only prevented ingestion jobs from executing on Feb 8th. Your existing metadata, lineage, and data catalog remain completely intact.

Technical Details

Root Cause: Unpinned dependency on setuptools. After an earlier dependency concern with the ruamel-yaml package, most dependencies for the datahub cli were pinned. However, setuptools was not.

Impact: All SQL-based sources that support profiling were impacted. Athena, BigQuery, ClickHouse, CockroachDB, Druid, Hive, IBM Db2, MariaDB, Microsoft SQL Server, MySQL, Oracle, PostgreSQL, Presto, Redshift, SAP HANA, Snowflake, SQLAlchemy Generic, Teradata, Trino, Unity Catalog, Vertica were impacted.

What We’ve Done:

For all existing recipes in managed instances, we have pinned the version of setuptools to be below 82.0.0. For all managed customers on v0.3.16.x and v0.3.15.x server versions, we have upgraded the default CLI version so all new recipes will also have setuptools pinned below 82.0.0. For customers below 0.3.15.x, or on customer-managed instances, we will reach out to you to coordinate a fix.

Workaround for all versions:

Add [“setuptools<82.0.0"] to Extra Pip Libraries

Who Needs to Take Action

Existing recipes in DataHub Cloud Managed Ingestion created before Feb 8th, 2026: no action required.

Action required if you run ingestion via:

  • CLI outside of Managed Ingestion
  • Managed Ingestion in self-hosted DataHub instances
  • Any customer-managed DataHub deployment

Remediation

The attached guide provides detailed instructions for three scenarios:

  1. CLI Users: Upgrade to DataHub CLI v1.3.1.7.post1 or re-install datahub with setuptools pinned: `pip install "setuptools<82.0.0" acryl-datahub`. v1.3.1.7.post1 is recommended for datahub versions 0.3.16.X or above.
  2. Cloud Instances (servers =0.3.16.X): Set CLI version to v1.3.1.7.post1 or add [“setuptools<82.0.0"] to Extra Pip Libraries
  3. Cloud Instances (servers <=0.3.15.X): Add [“setuptools<82.0.0"] to Extra Pip Libraries while we investigate compatibility between 0.3.15 and 1.4.0.2.
  4. Self-Hosted Instances: Set Helm value global.datahub.managed_ingestion.defaultCliVersion to v1.3.1.7.post1 or any of the recommended versions mentioned in Resolution above.

After applying remediation, verify with: python3 -m datahub version and run a test ingestion.

Next Steps

We are actively scanning for and remediating affected recipes, upgrading customer environments, and reviewing our dependency management processes to prevent similar issues.

Please review the attached remediation guide for step-by-step instructions. If you have questions or need assistance, contact our support team via our Support Portal.

Best regards,

The DataHub Team

TL;DR (Summary of Issue and Actions)

  • On Feb 8, 2026, ingestion jobs failed due to a breaking change in setuptools 82.0.0

  • No data was lost or corrupted

  • The issue only affected ingestion execution

  • DataHub Cloud Managed Ingestion has already been fixed

  • Action is required if you run ingestion via:

    • The DataHub CLI outside Managed Ingestion

    • Self-hosted or customer-managed DataHub

  • Fix by pinning setuptools<82.0.0 or upgrading to CLI v1.3.1.7.post1

Related to

Related articles