Free AWS Certified SysOps Administrator - Associate practice — 6 questions on Reliability and Business Continuity, with explanations. No sign-up.
Full 12-question mixed test →
Question 1 of 6 · Reliability and Business Continuity
A financial services company runs an RDS for PostgreSQL primary database in us-east-1. Leadership requires that a single-AZ hardware or network failure be resolved automatically with minimal downtime, and that a full regional disaster be recoverable with an RPO of 5 minutes by promoting a copy in eu-west-1. Which design meets BOTH requirements?
Multi-AZ provides automatic failover to a synchronous standby within the same region in roughly 60-120 seconds, satisfying the AZ-failure requirement, while a cross-region read replica typically lags by seconds to a few minutes and can be manually promoted to meet a 5-minute RPO during a full regional disaster.
Question 2 of 6 · Reliability and Business Continuity
An Auto Scaling group launches new EC2 instances behind an ALB target group. Application initialization via user data takes 8 minutes before the instance can pass the target group's health check (interval 30 seconds, unhealthy threshold 2). New instances are being terminated and replaced by the Auto Scaling group before initialization finishes. Which single change resolves this?
HealthCheckGracePeriod tells the Auto Scaling group to ignore EC2 and ELB health check results for the configured number of seconds after instance launch, giving the 8-minute initialization process time to complete before the ASG can evaluate and act on health status.
Question 3 of 6 · Reliability and Business Continuity
A company wants Route 53 failover routing to declare a primary endpoint unhealthy and shift traffic to a secondary endpoint only after approximately 90 seconds of continuous failure, using standard global health checker consensus. Which health check configuration achieves this?
Route 53 health checkers evaluate at the configured request interval and declare an endpoint unhealthy after consecutive failures equal to FailureThreshold, using majority consensus among global checker locations; a 30-second interval multiplied by a threshold of 3 yields the required approximately 90-second detection window.
Question 4 of 6 · Reliability and Business Continuity
A company runs critical EC2/EBS workloads in ap-southeast-1 and must be able to launch replacement resources in ap-southeast-2 within 15 minutes of a regional outage, using data no older than 1 hour (RPO). Which solution is the MOST operationally efficient way to meet this requirement?
AWS Backup centralizes policy-based backup scheduling and supports a native cross-region copy rule, enabling automated hourly recovery points in the DR region and a fast, repeatable restore process that meets both the 1-hour RPO and the 15-minute RTO target.
Question 5 of 6 · Reliability and Business Continuity
A team configured S3 Cross-Region Replication (CRR) from us-east-1 (source) to us-west-2 (destination) with versioning enabled on both buckets for disaster recovery. After issuing DELETE requests without specifying version IDs on several objects in the source bucket to reduce storage costs, the team observes that destination bucket storage usage does NOT decrease, and prior object versions still appear when listed with version-aware tools. Which explanation is correct?
By default, S3 CRR replicates delete markers created by non-versioned DELETE requests, which hides the current object version at the destination without removing any underlying version data; destination storage only decreases once noncurrent versions are explicitly expired via a lifecycle rule.
Question 6 of 6 · Reliability and Business Continuity
An Auto Scaling group behind an ALB scales out during traffic spikes. New instances take about 3 minutes to warm up an in-memory JVM cache, causing intermittent 502 errors immediately after instances are marked healthy and begin receiving full traffic. Which single change reduces these 502 errors WITHOUT slowing down failure detection for genuinely unhealthy targets?
Slow Start mode gradually increases the proportion of requests an ALB sends to a newly healthy target over the configured duration (30-900 seconds), preventing an immediate flood of traffic before the cache warm-up completes, and it does not affect how quickly the ALB detects genuinely unhealthy targets.
Ready for the real thing?
The full course: two full-length practice tests, video lessons for every exam domain, hands-on labs and detailed explanations.
undefined$34.99 with code 89CE4ED79F19FA855B6D — valid through Sep 23.