Designing Global High-Availability Bigtable Solutions
Short answer
C Implement a multi-cluster Bigtable instance across multiple regions with replication to ensure global high availability and scalability for time-series data.
Question
Key Takeaways
The main trap is prioritizing autoscaling over geographic distribution; true global high availability requires multi-region replication, not just scaling within a region.
This page explains why implementing a multi-cluster Bigtable instance across multiple regions is the correct design for global, high-availability time-series data platforms on Google Cloud.
Common mistake: Choosing Option B (Multi-cluster with autoscaling) is a common mistake because autoscaling addresses performance spikes, but it fails to address the 'global' and 'multi-region' availability requirements inherent in the scenario.
Expert Analysis
The scenario specifies a global, 24/7 platform requiring high availability for mission-critical operations and the ability to handle traffic spikes. The data is versioned time-series, which points directly to Google Cloud Bigtable as the optimal storage engine, eliminating AlloyDB (Option D) which is optimized for transactional PostgreSQL workloads rather than high-throughput time-series analytics.
Between the Bigtable options, Option A is insufficient because a single-cluster instance does not provide global reach or the high availability required for a 24/7 global service. Option B suggests a multi-cluster instance with autoscaling. While autoscaling helps with capacity, it does not inherently guarantee the low-latency global access or the specific regional redundancy needed for a "global" platform without explicitly defining the geographic distribution.
Option C is the correct choice because it explicitly states across multiple regions with replication. For a global application, deploying clusters in different regions allows users to connect to the nearest cluster, reducing latency. Multi-region replication ensures that if one region fails, the service remains available, satisfying the "highly available" and "mission-critical" requirements. As noted in community feedback, while autoscaling is beneficial, the architectural decision to span multiple regions is the primary driver for meeting the global HA requirement. Without multi-region placement, you cannot truly serve a global user base with high availability.
Key Insight: The question tests the distinction between scalability (handling load) and availability/geography (serving global users reliably). Autoscaling handles load, but multi-region deployment handles global availability and latency. Getting this wrong leads to designing a solution that may scale well locally but fails globally during regional outages.
Exam Strategy
When designing for 'global' availability, always prioritize geographic distribution (multi-region/multi-cluster) over local scaling features alone. Identify keywords like 'global', '24/7', and 'mission-critical' to trigger the need for cross-regional replication and failover capabilities, distinguishing them from simple throughput requirements.
Frequently Asked Questions
Why not use AlloyDB for time-series data?
Does autoscaling replace the need for multi-region deployment?
Community Votes
75% of anonymous learners picked answer B. Votes are pick records left by other test-takers — they are not the verified answer.
Discussion 4 contributions
Compiled from public community threads and reader submissions, lightly edited for clarity. Handles and vote tallies are shown as recorded, and comments reflect the views of individual contributors.
From public community threads 4
-
cachopo 👍 3 Answered CThis is the correct choice because Bigtable is the Google Cloud service designed specifically for high-throughput, versioned time-series data, and deploying it across multiple regions is the only way to satisfy the "Global" and "Mission-critical High Availability" requirements simultaneously.
-
yokoyan 👍 2 Answered BB is the correct answer.
-
AwesomeGCP 👍 1 Answered BThe correct answer is B. Implement a multi-cluster Bigtable instance with autoscaling enabled and optimal schema design. Why Option B is Correct High Availability & Zero Downtime: A multi-cluster Bigtable instance replicates data across clusters, providing high availability (up to 99.999% SLA depending on configuration) and automated failover for mission-critical, 24/7 global operations. Elasticity for Spikes: Combining multi-cluster replication with Autoscaling allows Bigtable to dynamically add and remove nodes based on incoming CPU utilization and storage load, gracefully handling sudden traffic surges without overpaying during off-peak times. Large-Scale Versioned Time-Series: Bigtable natively supports multi-versioned time-series data using cell timestamps and column families, provided the schema (row key design) is properly optimized to prevent write hot-spotting.
-
Alex0707 👍 1 Answered BAutoscaling is an important criteria according to the requirements and C is missing it
Reader comments & corrections
No reader corrections yet — spotted an error or have a note? Share it below.