An enterprise logistics company is migrating its fleet monitoring application to Compute Engine on Google Cloud. The system architecture has two key storage requirements:
1. A fully managed NoSQL database capable of ingesting high-throughput, low-latency time-series GPS location telemetry from tens of thousands of delivery vehicles.
2. Extremely high IOPS local block storage attached to processing virtual machines to serve as high-speed, temporary scratch space for processing incoming data streams, where data loss upon instance termination is acceptable.
Which TWO Google Cloud database and block storage options should the engineering team select? (Select TWO.)
- Cloud Bigtable for storing high-throughput, low-latency GPS telemetry dataCevap
- Local SSD attached to Compute Engine instances for temporary, high-IOPS stream transformation scratch disk spaceCevap
- CCloud SQL for PostgreSQL for storing high-throughput, low-latency GPS telemetry data
- DStandard Persistent Disk (pd-standard) attached to Spot VMs for high-IOPS temporary stream transformation scratch disk space
Cevap
The correct architecture requires selecting Cloud Bigtable for the time-series telemetry ingestion and Local SSD for high-performance temporary VM scratch space.
The solution requires Cloud Bigtable because it is engineered for high-volume, low-latency NoSQL time-series data such as vehicle GPS tracking. Local SSD is the ideal choice for ephemeral scratch space requiring maximum IOPS, as it is physically attached to the hypervisor host and yields superior performance for scratch workloads.
Adım Adım Çözüm
Anahtar Kavram
Selecting GCP Managed Databases and Persistent Block Storage Types based on Workload Characteristics