Running an Apache Iceberg Lakehouse With No Internet Connection
A practical guide to deploying an Iceberg lakehouse in air-gapped environments: component choices, artifact pipelines, identity without a cloud.
The archive
669 posts on Apache Iceberg, lakehouse architecture, data engineering and applied AI.
A practical guide to deploying an Iceberg lakehouse in air-gapped environments: component choices, artifact pipelines, identity without a cloud.
Autonomous materialized view management replaces quarterly review meetings with workload-driven scoring, and it's essential when AI agents generate.
Agents fail on raw lake data because business rules live in people's heads. Data products with semantic contracts fix this at the source.
Equality deletes made streaming upserts into Iceberg practical at the cost of read performance.
Agent reliability is a property of the stack the model sits on. Five layers with distinct owners and failure modes turn the agent is unreliable.
Fluss puts a columnar, indexed hot tier between Kafka and Iceberg. Here's what it changes structurally, what Kafka still does better, and how to benchmark.
A lakehouse cannot serve sub-second queries over seconds-old data. A hot tier in front solves it, with consequences for consistency, governance.
Commit conflicts multiply with writer count, and AI agents introduce unpredictable write patterns.
Jackson 3 changes everything: package names, unchecked exceptions, flipped defaults. Here's what breaks, why the engines are fine and your service isn't.
The catalog is the right attachment point for AI agents working against a lakehouse. Here's how to wire the official Polaris MCP Server and add the read.