Lab 4A — S3 lifecycle + cost

15 timed scenario questions (~24 min, ~96 sec each). Reinforces night-30-s3-lifecycle-cost.md and gsa-s3-lifecycle-map.md. Answer without peeking at night-30-quiz-answers.json first.

1. 1.Global Ski Atlas stores monthly Iceland pipeline output under globalskiatlas-backend-k8s-output/iceland/YYYY-MM/. Runs older than 90 days are read at most once per quarter; after one year they are pure archive. Which automation fits with minimal ops?

   A.Manually change each object's storage class in the console every month by prefixB.Enable S3 Transfer Acceleration on the bucket for all iceland/ pipeline PUT callsC.Lifecycle on iceland/ prefix: age-based moves to Standard-IA then Glacier classesD.Migrate all pipeline objects onto EBS gp3 volumes attached to the ECS Fargate task

2. 2.You create a lifecycle rule to transition objects from S3 Standard to Standard-IA after 1 day. Objects uploaded yesterday remain Standard today. What exam concept explains the delay?

   A.Intelligent-Tiering requires ninety days before any automatic tier move can occurB.Minimum storage duration — objects must stay in the current class at least 30 daysC.Glacier Vault Lock prevents all transitions until the configured legal hold expiresD.S3 lifecycle evaluation runs only on the first Sunday of each calendar month

3. 3.A new analytics prefix has unpredictable access — some objects hot for weeks, others cold for months. You want AWS to move tiers automatically without guessing Standard vs IA upfront. Which storage class?

   A.S3 Intelligent-Tiering with a small per-object monitoring and automation feeB.S3 Glacier Deep Archive applied immediately to every newly uploaded objectC.S3 One Zone-IA because it is always cheaper than Standard for every workloadD.EBS sc1 volumes attached to a Lambda function as a substitute for object storage

4. 4.Your study account needs an email when forecasted monthly spend exceeds 80% of a $25 cap. Which service sends that notification?

   A.AWS Cost and Usage Report delivered nightly into an S3 analytics bucketB.Amazon CloudWatch alarm on the S3 NumberOfObjects bucket metric aloneC.AWS Cost Explorer saved report that visualizes spend but sends no alertsD.AWS Budgets with ACTUAL or FORECASTED threshold email notifications

5. 5.Finance wants hourly line-item billing exports to a data lake for custom SQL in Athena. Which cost tool is purpose-built?

   A.S3 Storage Lens dashboard only — no line-item export path for Athena SQLB.AWS Cost and Usage Report (CUR) published to S3 for custom Athena analysisC.AWS Budgets with SNS fan-out to three SQS queues for billing line itemsD.Amazon Macie classification jobs run against the billing console export UI

6. 6.Reproducible pipeline artifacts can be rebuilt from source if one Availability Zone fails. You want the cheapest infrequent-access S3 tier and accept AZ-level risk. Which class?

   A.S3 One Zone-IA — lower cost than Standard-IA, data stored in a single AZB.S3 Standard with cross-Region replication configured to three backup RegionsC.S3 Glacier Deep Archive advertised with instant millisecond object retrievalD.Amazon EFS Standard spanning all Availability Zones in the Region for objects

7. 7.After lifecycle moves old compliance PDFs to Glacier Flexible Retrieval, auditors request a 200 GB restore. Storage cost dropped, but finance sees a new charge. What happened?

   A.Lifecycle automatically upgrades archived objects back to S3 Standard on every GETB.Glacier eliminates all retrieval fees — any new charge must be treated as a billing bugC.Glacier retrieval is billed separately — lifecycle lowers storage but not restore costD.AWS Budgets charges a fixed fee for each lifecycle transition event that S3 executes

8. 8.Multipart uploads to a pipeline bucket stall during network outages, leaving incomplete uploads that accrue storage charges. Which lifecycle element fixes this with no manual sweeper?

   A.S3 Object Lock in compliance mode retaining incomplete parts for seven daysB.Cross-Region Replication copying incomplete multipart uploads to a second bucketC.S3 Inventory report delivered weekly to SNS listing incomplete upload IDsD.AbortIncompleteMultipartUpload lifecycle rule that cleans up parts after N days

9. 9.You intended to archive iceland/2025-* prefixes but accidentally set the lifecycle filter to icland/ (typo). Six months later old pipeline data is still Standard. What failed?

   A.Glacier requires objects larger than 128 KiB — small GeoJSON files were skippedB.Lifecycle prefix filters are exact — the typo meant no objects matched the ruleC.Standard-IA cannot be targeted by lifecycle — only Glacier classes are allowedD.Cost Explorer must approve each transition before S3 is allowed to execute it

10. 10.The Iceland ECS task finished this morning and uploaded iceland/2026-07/ output. Downstream workers may re-read those objects all week. Which storage class should they remain in immediately after upload?

    A.S3 Standard — frequent access, millisecond retrieval, no IA retrieval surchargeB.S3 Glacier Deep Archive — lowest storage cost even for brand-new hot pipeline dataC.S3 One Zone-IA — always correct for any object younger than the 30-day minimumD.EC2 Instance Store — treated as the durable system of record for pipeline output

11. 11.A versioned bucket keeps overwritten wiki exports. You want lifecycle to move only older noncurrent versions to Standard-IA after 30 days. What must be true?

    A.Bucket must use S3 Object Lambda access points before versions can transitionB.S3 Transfer Acceleration must be enabled on the bucket for version transitionsC.Bucket versioning enabled and NoncurrentVersionTransitions defined in lifecycleD.AWS Backup plan must wrap the bucket — lifecycle cannot transition object versions

12. 12.Cost Explorer shows Amazon EC2-Other (NAT Gateway) dominates the study VPC bill while S3 storage is under $1. Which cost optimization aligns with Nights 8–10?

    A.Transition all S3 objects to Glacier Deep Archive immediately to eliminate NAT spendB.Add VPC gateway endpoints for S3 (and interface endpoints for ECR) to cut NAT egressC.Replace S3 with EBS io2 volumes for Iceland pipeline output to avoid NAT Gateway feesD.Disable S3 Block Public Access so clients bypass NAT Gateway for object downloads

13. 13.A junior engineer proposes a lifecycle expiration rule deleting iceland/ objects after 400 days — the only copy in one Region, no replication, no AWS Backup. What is the primary architecture risk?

    A.Expiration is free of charge — there is no meaningful risk if the objects are oldB.Glacier Flexible Retrieval cannot be expired — the lifecycle rule will always failC.Intelligent-Tiering blocks expiration until the monitoring fee invoice has been paidD.Permanent data loss if no replica — expiration is not DR; add replication or backup first

14. 14.Legal requires 7-year retention of rarely accessed audit logs with retrieval acceptable after half a day. Lowest storage cost is the priority. Which Glacier tier?

    A.S3 Glacier Deep Archive — lowest $/GB with retrieval measured in hoursB.S3 Glacier Instant Retrieval — millisecond access similar to Standard classC.S3 Standard-IA — cheapest option for every long-term archive compliance workloadD.S3 Intelligent-Tiering Frequent Access tier only for the full seven-year retention

15. 15.Night 30 applies lifecycle to iceland/ prefixes on globalskiatlas-backend-k8s-output. Where does durable Iceland pipeline output live before any transition — per Nights 10, 27, and 29?

    A.Attached EBS gp3 volume on the Fargate task — lifecycle moves EBS blocks into GlacierB.ElastiCache Redis — lifecycle flushes cache keys into S3 archive prefixes overnightC.S3 object keys under the pipeline bucket — EBS/Instance Store is scratch only on computeD.Aurora resort_stats table — S3 lifecycle rules apply automatically to SQL row storage