Prepare and test your skills

Prepare and test your skills

Free sample question with answers and detailed explanation below. The Professional Cloud Network Engineer practice bank has an additional 2,437 practice questions.Try 10 free realistic questions: instant score, no credit card, and no account required.
A global external Application Load Balancer uses one Compute Engine backend service for api.example.com, portal.example.com, and promo.example.com. Its Cloud Armor Enterprise policy uses Adaptive Protection and an evaluateAdaptiveProtectionAutoDeploy() rule with default thresholds.
A legitimate 2x surge on promo.example.com pushed load to 85%, triggering an alert (0.65 confidence, 0.008 impacted baseline) and auto-deploying a rule blocking valid users across all hosts. Preserving your baseline, you must:
Which set of actions should you take?
networksecurity.googleapis.com/dos_attack entries with confidence above 0.7, update the existing policy with --layer7-ddos-defense-auto-deploy-load-threshold=0.9 and --layer7-ddos-defense-auto-deploy-impacted-baseline-threshold=0.05, and set the evaluateAdaptiveProtectionAutoDeploy() rule to preview mode.Filtering networksecurity.googleapis.com/dos_attack logs in Cloud Monitoring reduces operator notification noise, while preview mode logs rule matches without dropping traffic. However, this combination misconfigures the impacted baseline threshold and disables automated DDoS mitigation.
--layer7-ddos-defense-auto-deploy-impacted-baseline-threshold parameter sets the maximum percentage of legitimate baseline traffic a suggested rule is allowed to affect (default 0.01, or 1%). Raising it to 0.05 (5%) relaxes the safety guardrail, permitting rules with five times more collateral damage—including the 0.008 event—rather than blocking them.evaluateAdaptiveProtectionAutoDeploy() in preview mode only logs what would have been blocked; it will not mitigate Layer 7 DDoS floods when backend load exceeds 90%.0.008), not increased.previewSecurityPolicy log fields without enforcing deny, throttle, or redirect actions.HTTP_HEADER_HOST granularity, traffic signatures remain evaluated against the entire backend service.You only wanted to audit alert confidence scores above 0.7 without enforcing automatic mitigation, and you intended to tolerate rules that affect up to 5% of baseline traffic during testing.
promo.example.com, and create a second new security policy with --layer7-ddos-defense-auto-deploy-load-threshold=0.9 and --layer7-ddos-defense-auto-deploy-confidence-threshold=0.7 for api.example.com and portal.example.com.Creating a new Google Cloud Armor security policy initializes a clean policy resource with no prior traffic history attached to it. While separating volatile workloads into a dedicated policy can reduce noisy alerts when workloads use separate backend services, creating new policies here breaks baseline continuity and fails to address a multi-host backend service.
api.example.com and portal.example.com, this approach aggregates traffic together and never reports the false positive to retrain the project-local machine learning (ML) models.BASELINE_TOO_RECENT status until sufficient traffic metadata is accumulated.HTTP_HEADER_HOST).The promo.example.com workload were already deployed on its own independent backend service with chronically unpredictable baseline swings, and you were willing to accept an hour-long re-training window while permanently disabling Adaptive Protection for that promotional backend.
--layer7-ddos-defense-auto-deploy-confidence-threshold=0.7 with add-layer7-ddos-defense-threshold-config (--detection-relative-to-baseline-qps=2.5), enable Cloud CDN on the backend service so Adaptive Protection baselines cached traffic, and move evaluateAdaptiveProtectionAutoDeploy() to a higher priority than explicit allow rules.This option attempts to pair legacy policy-level auto-deploy flags with granular threshold configs, cache traffic with Cloud CDN, and elevate the auto-deploy rule priority. Each of these actions violates core Google Cloud Armor constraints.
auto_deploy_config (--layer7-ddos-defense-auto-deploy-* flags) and threshold_config (add-layer7-ddos-defense-threshold-config) on the same security policy; configuring both causes an API error.evaluateAdaptiveProtectionAutoDeploy() at a higher priority (lower numerical value) than static allow rules can cause trusted clients or internal scanners to be inadvertently blocked by an automated signature.HTTP_HEADER_HOST granular traffic units nor reports false positive events to retrain the model.auto_deploy_config or threshold_config (which includes --auto-deploy-impacted-baseline-threshold and detection thresholds).evaluateAdaptiveProtectionAutoDeploy() should sit at a lower priority than explicit allow/deny rules.There is no valid scenario where combining auto_deploy_config and threshold_config succeeds; threshold_config alone must be used when configuring relative QPS or granular thresholds, while keeping static allow rules at higher priority.
add-layer7-ddos-defense-threshold-config to enable HTTP_HEADER_HOST granular traffic units, set --detection-relative-to-baseline-qps=2.5, --detection-load-threshold=0.9 at the backend service level, and --auto-deploy-impacted-baseline-threshold=0.005, and scope evaluateAdaptiveProtectionAutoDeploy() rules with request.headers['host'].Google Cloud Armor Adaptive Protection uses machine learning models to baseline Layer 7 traffic per backend service and detect HTTP floods. By updating the existing security policy with granular models, calibrated Layer 7 DDoS defense thresholds, host-scoped auto-deploy placeholder rules, and event error feedback, you preserve the learned baseline while eliminating collateral false positives.
--traffic-granularity-configs=type=HTTP_HEADER_HOST;enableEachUniqueValue=true turns each hostname into a distinct granular traffic unit. Because an auto-deployed rule still evaluates all requests matching the rule expression, combining evaluateAdaptiveProtectionAutoDeploy() with request.headers['host'] ensures a mitigation for one host never blocks legitimate users on api.example.com or portal.example.com.--detection-relative-to-baseline-qps to 2.5 prevents 2x surges from triggering detection, setting --detection-load-threshold=0.9 at the backend service level ensures automatic action only occurs above 90% load (instead of the 0.8 default), and lowering --auto-deploy-impacted-baseline-threshold to 0.005 prevents rules with a 0.008 (0.8%) baseline impact from auto-deploying.This approach uses native threshold configs, granular traffic units, host-scoped match expressions, and supervised model feedback on the active policy—solving false positives across shared backends without sacrificing real-time DDoS protection.
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