What Is Google Cloud Storage?

Google Cloud Storage (GCS) is Google Cloud Platform's object storage service, providing unlimited, globally consistent storage for structured and unstructured data. GCS stores data as objects in buckets, accessed via an HTTPS API compatible with both GCS-native clients and S3-compatible tooling. For Apache Iceberg lakehouses on GCP, GCS stores all data files and metadata, accessed using the gs:// URI scheme.

GCS provides strong read-after-write consistency for all operations globally — objects are immediately visible after upload, list operations reflect all objects, and deletes are immediately effective. This consistency model, combined with Google's global infrastructure, makes GCS a reliable foundation for Iceberg's atomic commit mechanism.

GCS Storage Classes

Like Amazon S3, GCS offers storage class tiers for cost optimization:

  • Standard: Most frequently accessed data. Highest cost, lowest access latency. Default for active lakehouse tables.
  • Nearline: Data accessed less than once per month. Lower storage cost, minimum 30-day storage duration.
  • Coldline: Data accessed less than once per quarter. Very low storage cost, minimum 90-day duration.
  • Archive: Long-term compliance data. Lowest cost, minimum 365-day duration, higher retrieval cost.
GCS Lakehouse Architecture diagram
Figure 1: GCS as GCP lakehouse storage — Iceberg on GCS with BigLake and Dataproc integration.

GCS and BigLake

Google BigLake extends BigQuery to query data stored in GCS — including Apache Iceberg tables. BigLake tables registered in a BigQuery project can be queried using BigQuery SQL, with GCS providing the actual storage. This enables organizations to use BigQuery as a query engine for their open Iceberg tables without moving data into BigQuery's native storage.

BigLake also provides fine-grained access control for GCS-stored Iceberg data: column-level security and row-level filters can be applied through BigLake policies, enforced regardless of which engine queries the table.

BigLake Iceberg on GCS diagram
Figure 2: BigLake enables BigQuery SQL on Iceberg tables stored in GCS.

Summary

Google Cloud Storage provides the same decoupled, scalable, durable object storage foundation for GCP-based data lakehouses that Amazon S3 provides on AWS. Its native GCP integration, BigLake support for Iceberg, and strong global consistency make it the natural storage choice for organizations building on Google Cloud. As with all cloud object storage services, GCS stores data in open Parquet format that any engine can read — maintaining the openness and portability of the lakehouse architecture.