AWS Big Data Blog

Learn how to set up Amazon Redshift Data Sharing between a producer and consumer cluster and use dbt Cloud to transform the shared data into analytics-ready models, with live, zero-copy access and no ETL between clusters.

Amazon Redshift now automatically resolves many concurrent writes to Apache Iceberg tables that don’t actually overlap, committing changes that previously failed and had to retry. Learn why these writes conflict, how the engine rebases non-overlapping ones across operation, partition, and file checks, and the throughput gains it delivers.

AWS Glue Data Quality is now available natively on Amazon EMR, bundled in AWS runtime for Apache Spark (emr-spark-8.1.0). Write the same DQDL rulesets you use in AWS Glue directly in your Amazon EMR Spark jobs. This post walks through two patterns: lightweight inline validation and an Apache Iceberg Write-Audit-Publish workflow with managed rulesets.

Learn how to build an enterprise-scale cross-account streaming pipeline with Amazon Kinesis Data Streams using the Kinesis Producer Library (KPL) and Kinesis Client Library (KCL) across three separate AWS accounts, with resource-based policies, cross-account IAM, and a shared customer managed AWS KMS key for maximum security isolation.

Troubleshooting a multi-service, MWAA-orchestrated ETL pipeline often means hunting across scattered Amazon CloudWatch log groups. This post shows how to centralize your ETL logs in Amazon OpenSearch Service and query them in plain language through an MCP server on Amazon Bedrock AgentCore, so you find root causes faster.

Learn how to pair DuckDB with AWS Glue 6.0 to run SQL-centric ETL on a single worker, reading Parquet from Amazon S3 and writing Apache Iceberg tables to Amazon S3 Tables. This post walks through a complete, runnable example and compares measured cost and runtime against an equivalent Apache Spark job on the same Glue runtime.

Amazon EMR 8.1.0 adds multi-catalog support through the RedirectingSessionCatalog, so you can query Iceberg, Delta Lake, Hudi, and Hive tables through a single catalog and join data across AWS accounts without copying it. This post shows how to put these capabilities into practice on Amazon EMR Serverless.

Amazon Redshift now supports Iceberg materialized views. Compute an aggregation once in Amazon Redshift and store the result as a standard Apache Iceberg table in Amazon S3, queryable by Amazon Athena, Apache Spark, Amazon SageMaker, and AWS Glue. This post covers use cases, incremental refresh, and a step-by-step getting-started guide.

Amazon Redshift automates much of its own tuning, but periodic operational reviews still pay off. Learn how the review_cluster tool in the open-source Amazon Redshift MCP server runs a full cluster diagnostic from one natural-language request, evaluating 12 diagnostic areas and returning prioritized, documentation-linked recommendations in minutes.

The KIP-848 consumer protocol in Apache Kafka 4.0 redesigns consumer group rebalancing to eliminate stop-the-world pauses. This post explains how the consumer protocol works on Amazon MSK, how to enable it on MSK Standard and Express brokers, and how to diagnose and resolve slow rebalancing issues.

Debugging Apache Airflow DAG failures across services like AWS Glue, Amazon EMR, and Amazon Athena is slow and manual. In this post, we show you how to build a custom Airflow plugin that integrates with Amazon Bedrock to automatically analyze DAG task failures and deliver on-demand root cause analysis on Amazon MWAA.

Announcing Spark Connect on Amazon EMR on EKS: build, test, and debug Spark applications from VS Code, PyCharm, Jupyter notebooks, Amazon SageMaker Unified Studio, or dbt, while running full-scale Spark operations on your existing Amazon EKS clusters.

Amazon Aurora PostgreSQL zero-ETL integration with Amazon SageMaker replicates your operational data to a lakehouse in near real time, without building custom ETL pipelines. Learn the architecture and change data capture mechanics, then set up the integration and query your data in Amazon SageMaker.

Amazon Redshift now supports evolving Apache Iceberg table schemas and partition layouts through ALTER statements, with no data rewrites or pipeline rebuilds. In this final post of the series, you rename, add, drop, and widen columns, evolve partitions, and create AWS Lake Formation resource links for governed cross-engine access to Amazon S3 Tables.

Amazon SageMaker Unified Studio now supports custom permissions boundaries for the IAM roles its Tooling blueprint creates. Learn how to create a permissions boundary that restricts AI agent capabilities, configure it on the Tooling blueprint with the AWS CLI, and validate that every provisioned role carries the boundary.

Learn how Delivery Hero rebuilt its ad measurement pipeline from hourly batch processing to real time on Amazon Managed Service for Apache Flink, cutting the gap between an event and its recording from 61 minutes to 1.2 seconds and reducing monthly operational costs by about 57%.

Configuring VPC networking per project across a SageMaker Unified Studio domain creates inconsistent, hard-to-audit networks. This post shows administrators how to configure domain-level VPC networking once, so every new project automatically inherits consistent, private network isolation, then update existing projects and validate connectivity.

Amazon EMR on EC2 now supports Spark Connect, so you can develop and debug PySpark interactively from Amazon SageMaker Unified Studio Data Notebooks or your own IDE while Spark runs on your cluster. This post shows you how to get started from both a Data Notebook and a local IDE.

In Part 2 of this series, sign in as a data consumer, subscribe to enriched unstructured data assets in Amazon SageMaker Catalog, and query them using natural language through a no-code Amazon Bedrock chat agent and Amazon Bedrock model inference.

As consumer groups on Amazon MSK scale to thousands of members, the metadata record Kafka writes during rebalances can exceed the 1 MB limit and stall the group. Learn how to estimate metadata size, raise the topic-level limit safely, plan capacity, and apply complementary strategies for scaling large consumer groups.