Control Key Data Usage for Multi-Region Clusters with Data Localization Restrictions

MediumLevel 1Manual reviewcockroachdb

Description

In self-hosted CockroachDB multi-region clusters, organizations must ensure that Personally Identifiable Information (PII) or other regulated data is not used as part of primary key constraints or secondary indexes. In CockroachDB, primary keys define the primary index, and primary key columns are implicitly included in all secondary indexes. When a column is included in a primary key or index, its values are: 1.

How it is checked

Audit method: Manual review

This recommendation cannot be checked automatically. A scan lists it with these instructions when you pass --include-manual.

If the cluster uses multi-region features (SHOW REGIONS FROM DATABASE; REGIONAL BY ROW / GLOBAL tables), identify tables holding PII subject to data-localization rules and review SHOW CREATE TABLE for each: PII must not be used in primary keys or secondary index keys (key data is replicated to meta ranges in all regions). Single-region clusters are not applicable.

How it is fixed

Fix method: Guidance or infrastructure patch

1. Replace PII-based keys with surrogate keys Refactor schemas so that primary keys use surrogate identifiers instead of PII. CockroachDB recommends: - Using UUID-typed columns with default values generated by gen_random_uuid() for single-column primary keys. - Using composite primary keys where the leading column is a well-distributed, non-sensitive identifier. Ensure PII columns are stored as regular, non-indexed fields where possible.

Properties

Operating systems: Any
Tags: cockroachdbdata-protection
Reference: CIS CockroachDB v25.x Benchmark v1.0.0 - 5.3 Control Key Data Usage for Multi-Region Clusters with Data Localization Restrictions