Data & AI

Apache Iceberg Metadata Architecture

Data & AI Pipelines diagram · made in Skemio

Tree-structured diagram of Apache Iceberg metadata showing Catalog pointing to Metadata File, pointing to Manifest List, referencing Manifest Files and Parquet files.
Tree-structured diagram of Apache Iceberg metadata showing Catalog pointing to Metadata File, pointing to Manifest List, referencing Manifest Files and Parquet files.

What it shows

  • Iceberg catalog stores an atomic pointer to the active metadata file.
  • Manifest lists track snapshots and partition summaries without file scans.
  • Manifest files hold file-level metrics enabling rapid partition pruning.

The takeaway

Iceberg isolates queries from raw file trees using hierarchical, atomic snapshot metadata.

More data & ai pipelines diagrams