Newsletters




Redpanda Shadowing Strengthens Disaster Recovery and Data Resilience in Redpanda Streaming 25.3


Redpanda, the real-time data platform for the agentic enterprise, is unveiling Shadowing, a new disaster recovery (DR) feature available in the upcoming Redpanda Streaming version 25.3.

According to the company, the capability provides an active–passive replication model that mirrors data from a primary cluster to a “shadow” cluster in a separate region, enabling rapid recovery and continuity in the event of regional cloud failures such as the recent AWS US-East-1 outage.

“Shadowing brings full enterprise-grade resilience to Redpanda Streaming,” said Tyler Akudau, CTO at Redpanda. “It’s about giving our customers the ability to design recovery architectures that fit their risk tolerance and compliance needs without unnecessary complexity. For customers who simply can’t afford to go dark, this capability is critical.”

Redpanda Streaming’s Shadowing capability eliminates complexity and risk by offering a built-in capability that pulls data and metadata directly into the DR cluster, preserving offsets, replicating schemas, and maintaining an identical replica in the secondary cluster.

If a region-wide outage occurs, operations can be redirected to the replica with minimal interruption and no loss of ordering or state, the company said.

In addition to Shadowing, this release introduces the beta of Cloud Topics and Serverless on Google Cloud Platform, along with key enhancements that expand Apache Iceberg integration and add Microsoft SQL Server CDC Connector.

Together, these updates make Redpanda even more flexible and accessible for teams building real-time and AI applications, the company said. Features include:

  • Cloud Topics Beta: A new topic class that replicates data directly to cloud object storage, such as Amazon S3, while copying only lightweight metadata across network zones to optimize latency-cost trade-offs and dramatically reduce cross-zone network traffic for high-throughput streaming workloads.
  • Apache Iceberg Topics for Google BigLake metastore: Redpanda’s native Iceberg integration now automatically registers stream-backed tables to Google BigLake metastore and makes them available for supported query engines like BigQuery.
  • Microsoft SQL Server CDC Connector for Redpanda Connect: Native change data capture (CDC) support for SQL Server, enabling seamless event streaming from legacy databases into Redpanda pipelines.
  • Serverless Beta on Google Cloud Platform: Redpanda’s fully managed serverless offering will soon become  available in Google Cloud, allowing users to instantly spin clusters up and down, and automatically adapt to changing workloads.

For more information about Redpanda 25.3, visit www.redpanda.com.


Sponsors