Skip to content
LakeBench
ProblemsCommunityPricing
Sign inStart practicing
Back
  1. Home
  2. Interview prep
  3. Regions, zones and data residency

Cloud · Cloud Architecture Basics

Regions, zones and data residency

Easycloud-43
cloud-architectureregionsdata-residencycostdisaster-recovery

Question

How do regions and multi-regions affect a data platform design?

Solution

Cloud regions define the physical geographical locations of data centers, dictating data residency compliance, query performance, and network egress costs. Regulations like GDPR require customer data to remain within specific geographic boundaries, while analytical platforms like BigQuery enforce location boundaries that prevent joining datasets located in different regions. Multi-region storage configurations provide high availability and regional durability, but cross-region data transfers introduce network egress fees that must be managed.

Data residency and analytical locality

Placing cloud storage in the wrong region can violate international privacy laws and cause runtime query errors.

EU Region (Germany)        US Region (Virginia)
     |                             |
BigQuery Dataset (EU)      BigQuery Dataset (US)
     \                             /
      X--- Cannot JOIN Directly --X  (Must replicate or copy data first)

Architectural decisions must account for physical regional boundaries:

  • Legal compliance and data residency regulations, such as GDPR in Europe or specific financial regulations in India and Canada, mandate that customer data never physically leaves designated national borders.
  • Analytics engines enforce strict regional boundaries. In Google Cloud BigQuery, you cannot run a single SQL query that joins a table in the us-central1 region with a table in the europe-west1 region without explicitly replicating or exporting the data.
  • Colocate your processing compute with your storage. Running Apache Spark compute workers in one region while reading terabytes of Parquet files from an object store bucket in another region slows execution down and incurs massive cross-region egress expenses.

Availability and disaster resilience

Teams choose between regional and multi-region deployment models based on risk tolerance:

  • Multi-region configurations (such as BigQuery US multi-region or GCS Dual-Region buckets) distribute data copies across geographically separated metropolitan areas within a continent, surviving data center outages.
  • A cross-region disaster recovery strategy replicates critical analytical datasets asynchronously to a secondary region, providing operational continuity if a primary region experiences an extended outage.
  • Network egress fees are charged whenever bytes traverse regional boundaries, meaning continuous cross-region replication must be reserved for business-critical datasets.
PreviousNext