Prepare and test your skills

Prepare and test your skills

Free sample question with answers and detailed explanation below. The Professional Cloud Database Engineer practice bank has an additional 769 practice questions.Try 10 free realistic questions: instant score, no credit card, and no account required.
A company is deploying a 4 TB transactional database on an N2 Compute Engine VM with 8 vCPUs and needs to select a block storage option that satisfies strict technical and budgetary requirements:
Which storage type should the company choose?
Local SSD is high-performance, physically attached ephemeral block storage hosted directly on the physical server running the Compute Engine VM. By avoiding network hops, Local SSDs deliver ultra-low latency and massive IOPS compared to network-attached Persistent Disks.
Despite easily surpassing the latency, IOPS, and throughput thresholds, Local SSD violates the durability and backup requirements:
The storage were intended strictly for ephemeral scratch space, hot analytics caching, or temporary database workspace (such as SQL Server tempdb) where data loss upon VM termination is acceptable.
Standard Persistent Disk (pd-standard) is an economical network block storage option backed by hard disk drives (HDD). It is engineered primarily for large, sequential read and write operations such as batch processing, streaming analytics, or infrequent archival access.
While pd-standard offers the lowest cost per GiB among Persistent Disk types, it fails multiple technical requirements for a transactional database:
pd-standard does not provide the sub-millisecond latency offered by SSD-backed Persistent Disks.pd-balanced and pd-ssd, both IOPS and throughput are half-duplex on pd-standard. Combined read and write bandwidth share a single ceiling, and maximum write throughput is capped at 400 MiB/s (or 200 MiB/s for regional disks), causing read and write traffic to compete for bandwidth.pd-standard volumes attached to the VM.The workload involved latency-insensitive batch processing, large sequential table scans, or storing >10 TB of archival data where minimizing storage cost outweighs random I/O performance.
SSD Persistent Disk (pd-ssd) is a high-performance, SSD-backed network block storage volume designed for enterprise applications and IOPS-intensive databases that require low latency and higher baseline IOPS per GiB than Balanced Persistent Disk.
Although pd-ssd satisfies all durability, latency, IOPS, and throughput specifications, it fails the budgetary constraint of minimizing monthly storage expenses:
pd-ssd and pd-balanced is capped at 15,000 IOPS and 800 MiB/s of independent read and write throughput.pd-balanced delivers—paying the higher per-GiB rate for pd-ssd yields no performance benefit under this VM shape.pd-balanced and pd-standard.The database were deployed on a larger VM (such as 16+ vCPUs, where pd-ssd unlocks up to 25,000 to 100,000 IOPS compared to lower caps on pd-balanced) or if the disk capacity were much smaller (e.g., 500 GiB) and needed higher IOPS per GiB than pd-balanced could supply.
Balanced Persistent Disk (pd-balanced) is a solid-state drive (SSD) network block storage option designed to provide durable, predictable performance at a lower price point than higher-tier SSD disks. It stores data redundantly across Google Cloud infrastructure and supports standard data protection features such as snapshots and disk cloning.
pd-balanced persists independently of the VM lifecycle and supports scheduled snapshots for backup and recovery.pd-balanced is engineered for sub-millisecond latencies and shares a single half-duplex IOPS limit between reads and writes. On an 8-vCPU N2 VM, pd-balanced supports up to 15,000 combined IOPS, which comfortably accommodates the 12,000 IOPS workload.pd-balanced are full-duplex (independent of each other). An 8-vCPU N2 instance supports up to 800 MiB/s for reads and 800 MiB/s for writes simultaneously, easily exceeding the 300 MiB/s requirement.pd-balanced costs significantly less per GiB than pd-ssd or pd-extreme, avoiding unnecessary expenditure.On an 8-vCPU N2 instance, both pd-balanced and pd-ssd share the exact same per-instance ceiling of 15,000 IOPS and 800 MiB/s read/write throughput. Because a 4 TB pd-balanced volume satisfies the 12,000 IOPS and 300 MiB/s requirements while costing less per GiB than pd-ssd, pd-balanced delivers the optimal balance of performance, durability, and budget efficiency.
Answer them yourself and get your score the moment you finish. No account needed.
Full access includes
Keep the momentum going with these hand-picked practice scenarios
Last updated Validated