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 runs a global real-time application backed by Cloud Bigtable with application servers in us-central1 and europe-west1. The team deploys Bigtable clusters in us-central1, europe-west1, and us-east1, dedicating the us-east1 cluster to heavy nightly batch analytics jobs. The architecture must meet these requirements:
Which Bigtable app profile configuration should the team implement?
In Cloud Bigtable, an application profile (app profile) defines the routing policy that controls which clusters handle incoming client requests. Cluster group routing is a specialized form of multi-cluster routing that limits traffic and automatic failover to an explicit subset of clusters within an instance. Meanwhile, single-cluster routing pins an app profile's requests exclusively to one designated cluster.
us-central1 and europe-west1 automatically send requests to their nearest serving cluster, minimizing network round-trip latency.us-central1 or europe-west1 becomes unavailable or overloaded, Bigtable automatically fails over incoming real-time requests to the remaining healthy cluster in the restricted cluster group.us-central1 and europe-west1 excludes us-east1 from receiving live traffic or failover traffic. Simultaneously, pointing the batch analytics job to a single-cluster routing app profile on us-east1 isolates heavy analytical reads from live user traffic.us-central1 or europe-west1 are automatically replicated to us-east1 so batch jobs always analyze up-to-date data.us-east1 only handles replication and batch reads, its node count and autoscaling thresholds can be tuned independently.This combination provides the geographic redundancy and automatic load balancing of multi-cluster routing while preserving strict workload isolation for batch analytics. Note that because multi-cluster routing provides eventual consistency, single-row transactions cannot be enabled on the real-time app profile.
Any cluster routing is the default form of multi-cluster routing in Cloud Bigtable. It makes every cluster in the Bigtable instance eligible to receive application requests and act as an automatic failover target, routing each request to the geographically nearest healthy cluster.
us-east1 is part of the same Bigtable instance, any cluster routing includes us-east1 in its pool of eligible failover targets. If us-central1 experiences an outage or transient latency spike, Bigtable routes North American real-time requests to us-east1 instead of europe-west1 because us-east1 is geographically closer.us-east1 batch analytics cluster during a failure or maintenance event can exhaust its CPU and degrade both real-time serving latency and nightly analytics jobs.All three clusters in us-central1, europe-west1, and us-east1 were dedicated to serving real-time application traffic and none needed to be reserved exclusively for isolated batch analytics workloads.
Row-affinity routing (also known as sticky routing) is a multi-cluster routing option that deterministically maps single-row read and write requests to a specific cluster based on the request's row key. Its primary purpose is to increase the rate of read-your-writes consistency in replicated Bigtable instances.
us-east1 remains part of the active routing pool, causing live single-row requests to be directed to the dedicated batch analytics cluster.us-central1 or europe-west1 can be sent across regions or continents, increasing user-facing latency.MutateRow or ReadRows with a single key; it is not enforced for multi-row scans or SQL ExecuteQuery calls.The application required a higher rate of read-your-writes consistency for single-row operations across multiple clusters within the same region and did not need to exclude a cluster or strictly optimize routing by client proximity.
Even when using row-affinity routing, Bigtable recommends combining it with cluster groups (--restrict-to) whenever you need predictable failover order or cluster isolation.
Single-cluster routing directs every request sent through an app profile to one explicitly specified cluster in a Bigtable instance. While data continues to replicate asynchronously to all other clusters in the instance, client traffic never leaves the assigned cluster unless the app profile is modified.
us-central1 cluster or region experiences an outage, requests using the us-central1 app profile will fail until an administrator manually updates the app profile to point to europe-west1.gcloud CLI.The real-time application required single-row transactions or strict read-your-writes consistency, or if the team explicitly wanted to prevent automatic cross-region failover to avoid cross-region network egress costs and latency spikes.
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