An operations engineer is managing a Cloud Bigtable cluster that processes high-throughput time-series data from IoT sensors. During daily peak ingestion windows, application latency increases significantly, and Cloud Monitoring indicates that CPU utilization on the Bigtable cluster repeatedly exceeds 80%. The engineer must restore acceptable read/write performance while maintaining cost efficiency and operational monitoring. Which TWO actions should the engineer take to address this issue?
- Increase the number of nodes in the Cloud Bigtable cluster or configure a programmatic autoscaling policy based on target CPU utilization.Answer
- BMigrate the cluster storage tier from SSD to HDD to increase read/write IOPS throughput during peak load.
- Create a Cloud Monitoring alerting policy tracking the node CPU utilization metric to proactively trigger capacity adjustments before performance degrades.Answer
- DExport the time-series data to a Cloud SQL for MySQL instance to handle high-volume write operations with lower latency.
Answer
The engineer should scale out the cluster by increasing node count or enabling autoscaling, and set up Cloud Monitoring alerts on node CPU utilization.
Scaling out node capacity directly mitigates CPU bottlenecks in Cloud Bigtable because processing capacity scales linearly with node count. Combining this with Cloud Monitoring alerting allows administrators to proactively manage workload spikes and maintain low latency.
Step-by-Step Solution
Key Concept
Cloud Bigtable Operational Performance Management and Monitoring