Design cost-optimized storage solutions

Oakridge Ledger runs an Amazon RDS for MySQL instance named ledger-prod with 200 GiB of gp3 storage. The database grows by an unpredictable 5 to 40 GiB per month depending on customer onboarding. Last quarter the instance ran out of space, and the operations team responded by manually resizing the storage to 1,500 GiB to avoid a repeat. The company wants to pay only for the storage it uses while never running out of space again, with minimal operational effort.

What should a solutions architect recommend?

1. AKeep the 1,500 GiB allocation and create a CloudWatch alarm on FreeStorageSpace to warn the team at 10 percent free.An alarm helps the team react, but it does nothing about paying for roughly 1,300 GiB of unused provisioned storage every month.
2. BConvert the instance to Amazon RDS Multi-AZ so that storage is spread across two Availability Zones.Multi-AZ provides a standby replica for availability; it doubles the storage charge and does not grow the volume automatically.
3. CMove the database to Amazon DynamoDB so that storage grows automatically.DynamoDB does scale storage automatically, but re-platforming a relational MySQL application to a NoSQL key-value store is a major rewrite and far from minimal effort.
4. DReduce the allocated storage to a right-sized value and enable RDS storage autoscaling with a maximum storage threshold.RDS storage autoscaling increases the allocated storage automatically when free space falls below a threshold, so the instance can start at a size close to actual usage, grow with demand, and stay capped at a maximum the company controls.

Pinecrest Learning hosts 80 TB of recorded lecture videos in an S3 bucket named pinecrest-lectures. Some videos are watched thousands of times a month, others are not opened for a year, and the popularity of any given video changes without warning when instructors update their course lists. Objects average 900 MB. The company wants to reduce storage cost automatically without any change in retrieval latency and without building its own access-tracking logic.

Which S3 storage class should a solutions architect recommend?

1. AS3 Standard-Infrequent AccessStandard-IA charges a per-GB retrieval fee on every read, so the heavily watched videos would cost more than they do in S3 Standard.
2. BS3 Intelligent-TieringIntelligent-Tiering monitors access per object and moves objects between frequent, infrequent, and archive-instant tiers automatically with no retrieval fees and no change in millisecond latency, which fits an unpredictable access pattern with large objects where the small monitoring charge is negligible.
3. CS3 Glacier Flexible RetrievalGlacier Flexible Retrieval requires a restore that takes minutes to hours before an object can be read, which breaks the requirement for unchanged retrieval latency.
4. DS3 One Zone-Infrequent AccessOne Zone-IA stores data in a single Availability Zone with lower durability guarantees and still charges per-GB retrieval fees, so it is neither safe for a primary video library nor cheaper for popular objects.

Tidewell Pharmaceuticals stores laboratory instrument exports in a versioned S3 bucket named tidewell-lab-exports. A cost review shows that 35 percent of the billed storage consists of noncurrent object versions that are more than 30 days old, and another 12 percent consists of parts from multipart uploads that were started by an unreliable upload script and never completed. The company must keep the current version of every object indefinitely and wants to eliminate the wasted storage automatically.

Which two actions should a solutions architect take? Each correct answer presents part of the solution.

1. ASuspend versioning on the bucket so that old versions are removed.Suspending versioning stops new versions from being created but does not delete existing noncurrent versions, and it removes the protection the company relies on for current objects.
2. BAdd a lifecycle rule with a NoncurrentVersionExpiration action that permanently deletes noncurrent versions 30 days after they become noncurrent.NoncurrentVersionExpiration targets only previous versions, so the current version of every object is retained while stale versions are removed automatically after 30 days.
3. CAdd a lifecycle rule with an Expiration action that deletes all objects 30 days after creation.An Expiration action on a versioned bucket places a delete marker on the current version, which hides the object and violates the requirement to keep current versions indefinitely.
4. DAdd a lifecycle rule with an AbortIncompleteMultipartUpload action that removes incomplete uploads after 7 days.Incomplete multipart upload parts are billed as storage until they are aborted; this lifecycle action cleans them up automatically after the specified number of days.
5. EEnable S3 Object Lock in compliance mode on the bucket.Object Lock prevents deletion for a retention period, which would make the wasted versions and parts harder to remove, not easier.

Halverson Radiology must retain 400 TB of diagnostic images for 10 years. After the first 90 days an image is almost never accessed, but when a physician does request one, the viewing application must display it within a few seconds. The company currently keeps everything in S3 Standard and wants to lower the cost of the images older than 90 days.

Which storage class should a lifecycle rule transition the images to after 90 days?

1. AS3 Glacier Deep ArchiveDeep Archive is the lowest-cost class but standard retrievals take up to 12 hours, so a physician could not view an image within seconds.
2. BS3 Glacier Flexible RetrievalFlexible Retrieval requires a restore request that takes minutes with expedited retrieval or hours with standard retrieval, which does not meet a seconds-level access requirement.
3. CS3 Standard-Infrequent AccessStandard-IA provides millisecond access but is priced for data accessed about once a month; for data accessed about once a year, Glacier Instant Retrieval stores the same data at a substantially lower per-GB price.
4. DS3 Glacier Instant RetrievalGlacier Instant Retrieval offers millisecond, synchronous access like S3 Standard at archive-level storage prices, and is designed exactly for long-lived data accessed roughly once a quarter or less that still needs immediate retrieval.