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.
Your organization manages a multi-database platform on Google Cloud and must adjust capacity before a seasonal traffic surge:
db-g1-small and needs production SLA coverage alongside higher disk throughput and IOPS.N2 primary instance and an N2 read pool, but vertically scaling the N2 read pool fails due to regional hardware capacity constraints.Which scaling strategy meets all of these technical and availability requirements?
This strategy applies service-specific vertical scaling and horizontal scaling rules across Cloud SQL, AlloyDB, and Bigtable to satisfy performance targets while protecting service availability and SLA coverage. Each database service handles machine shape changes, hardware capacity pools, and memory allocation differently, so matching the scaling mechanism to the underlying architecture prevents bottlenecks and downtime.
db-g1-small instance to a dedicated-core machine type restores Cloud SQL SLA coverage, because shared-core tiers are excluded from production SLAs. Increasing both vCPUs and storage capacity directly raises disk throughput and IOPS, since Cloud SQL disk performance scales with both disk size and vCPU count.N2 capacity constraint. AlloyDB provisions replacement VMs, shifts traffic while preserving the read pool IP address, and leaves the primary instance untouched without restart.Unlike approaches that blindly scale every tier vertically or modify the AlloyDB primary unnecessarily, this strategy respects the exact resource boundaries of each database engine. It resolves SLA gaps in Cloud SQL, works around regional compute shortages in AlloyDB with minimal disruption, and prevents cache thrashing in Bigtable.
This option correctly upgrades Cloud SQL to a dedicated-core machine type, but it forces an unnecessary machine series change on the AlloyDB primary instance and relies on vertical scaling in Bigtable's in-memory tier to expand cache capacity.
N2 to C4A triggers an unneeded replacement swap that briefly drops active write connections and introduces risk if C4A capacity is constrained for the high-availability standby node.The Bigtable cluster only needed higher read throughput for a small dataset that already fit comfortably in existing node memory, and the AlloyDB primary instance itself was CPU-bound and unable to scale within the N2 series.
When sizing Bigtable's in-memory tier, administrators often assume vertical node scaling adds memory alongside throughput. In practice, horizontal node addition is the only way to increase total cluster memory capacity in Bigtable.
This option correctly scales Bigtable horizontally to expand memory and throughput, but it leaves Cloud SQL on a shared-core db-g1-small shape while expanding storage alone, and it attempts to work around the AlloyDB capacity error by pinning read pool nodes to a specific zone.
db-f1-micro and db-g1-small) are strictly excluded from the Cloud SQL SLA and are intended only for low-cost development or testing. In addition, Cloud SQL disk throughput and IOPS depend on both disk size and the number of vCPUs, so increasing storage alone on a shared-core instance cannot deliver high production I/O performance.N2 capacity constraint.The Cloud SQL database were a non-production workload without SLA requirements, and Cloud SQL allowed independent zonal node pinning for AlloyDB.
In Cloud SQL, network and disk throughput are tightly coupled to vCPU allocation as well as storage volume size. Vertical scaling for I/O-intensive workloads requires upgrading both compute and storage rather than expanding disk capacity in isolation.
This option attempts to switch Cloud SQL to Enterprise Plus while keeping db-g1-small, add a second primary instance inside the AlloyDB cluster, and vertically scale Bigtable in-memory nodes to their maximum per-node throughput limit.
N2, C4A, C4, and C4D series; shared-core machine types like db-g1-small only exist in Enterprise edition and do not support Google Cloud Hyperdisk Balanced or production SLAs.C4, C4A, C4D (Enterprise Plus) and N4 (Enterprise), not shared-core shapes.No such configuration is valid in Google Cloud because shared-core shapes cannot run on Enterprise Plus, AlloyDB clusters disallow multiple primary instances, and Bigtable vertical scaling never adds RAM.
When migrating or upgrading Cloud SQL instances to Enterprise Plus for higher machine limits and data cache, you must select a supported dedicated-core series such as N2 or C4A, which automatically pairs with SSD or Hyperdisk Balanced storage.
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