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 fintech company is deploying a payment ledger on Cloud Spanner using a multi-region instance configuration (nam3) with SSD storage. You must determine the minimum manual compute capacity needed to meet these production requirements:
Which compute capacity should you provision for the Cloud Spanner instance?
A 3-node (3,000 PU) Cloud Spanner instance distributes three server tasks per zone across the multi-region topology. It provides 30 TiB of total storage allotment (10 TiB per node) and an approximate theoretical maximum of 8,100 standard writes per second (3 × 2,700) at 100% CPU utilization.
4,800 / 8,100). This exceeds both the 45% high-priority CPU target (3,645 writes/sec max) and the 50% total CPU target (4,050 writes/sec max) recommended for multi-region configurations.Peak sustained write load were 4,050 writes per second or less (within the 50% total CPU threshold) on a 25 TiB database, or if the workload used throughput-optimized writes to achieve higher write throughput per node.
In Cloud Spanner, 1 node is equivalent to 1,000 processing units (PUs), so a 2-node instance provides 2,000 PUs of compute capacity replicated across every zone in the multi-region topology. At 100% CPU utilization under optimal conditions, each node in a nam3 multi-region SSD configuration supports approximately 2,700 writes per second (for 1 KB rows) and 10 TiB of storage.
2 × 1,350), and 2,430 writes per second at the 45% high-priority CPU target.The database stored 20 TiB or less and used a regional configuration (which supports up to 3,500 writes/sec per node at 70% total CPU, or 2,450 writes/sec per node), or if peak write throughput were 2,400 writes/sec or less on a dataset under 20 TiB.
Provisioning 4 nodes (equivalent to 4,000 processing units) allocates four dedicated server tasks per zone across the nam3 multi-region topology. Each node provides 10 TiB of storage capacity and approximately 2,700 peak standard writes per second (at 100% CPU for 1 KB rows), giving the 4-node instance 40 TiB of storage capacity and 10,800 writes/sec of raw peak write capacity.
1,350 writes/sec (4,800 / 1,350 = 3.56 nodes). At 45% CPU, each node sustains 1,215 writes/sec (4,800 / 1,215 = 3.95 nodes). Both thresholds round up to 4 nodes, which supports up to 4,860 writes/sec at 45% CPU and 5,400 writes/sec at 50% CPU.4 × 10 TiB), easily accommodating the 25 TiB dataset while leaving 15 TiB for future growth.Unlike 2 or 3 nodes, which violate storage limits or multi-region failover CPU guidelines, and unlike 5 nodes, which wastes compute budget, 4 nodes precisely balances throughput, failover safety, and storage requirements.
A 5-node (5,000 PU) Cloud Spanner instance allocates 50 TiB of total database storage capacity and roughly 13,500 writes per second of theoretical maximum write capacity (at 100% CPU) in a nam3 multi-region SSD configuration. While this configuration easily meets both the storage and throughput targets, it exceeds the minimum capacity required.
25 TiB / 5 TiB = 5 nodes) instead of Cloud Spanner's 10 TiB per node limit.5 × 10 TiB), double the initial 25 TiB dataset.4,800 / 13,500), well below even the conservative 45% multi-region threshold.Peak sustained write load were between 4,861 and 5,400 writes per second (strictly enforcing the 45% high-priority CPU target), if the dataset exceeded 40 TiB (up to 50 TiB), or if the target database were Cloud Bigtable with SSD storage and a 5 TB per-node storage limit.
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