Study storage tiering, lifecycle, backup, migration, and transfer-cost decisions for the SAA-C03 cost-optimized storage objective.
Cost-optimized storage questions on SAA-C03 are about aligning storage class, retention, migration method, and access pattern to the workload. AWS is not asking you to memorize every price point. It is asking whether you can avoid paying for premium storage or transfer methods when the workload does not need them.
The current exam guide points to access options such as Requester Pays, cost-management tools, storage services, backup strategies, block-storage options, lifecycle design, hybrid storage options, storage tiering, storage access patterns, migration paths, and storage auto scaling.
| Requirement | Strongest first fit | Why |
|---|---|---|
| Infrequently accessed S3 data with retrieval still needed | S3 Standard-IA or Intelligent-Tiering depending pattern | Reduces cost without moving straight to archival |
| Long-term archival with slow retrieval tolerance | Glacier storage classes | Strongest low-cost archival path |
| Data with lifecycle transition path | S3 lifecycle rules | Automates cost reduction over time |
| Shared Linux file storage for many clients | EFS only when the shared file semantics are needed | File storage is usually costlier than S3 for object workloads |
| Boot volume or low-latency block requirement | Right-sized EBS volume family | Match SSD or HDD economics to the workload |
| Large-scale storage transfer with managed sync behavior | DataSync or Storage Gateway depending environment | Better fit than ad hoc transfer scripts |
SAA-C03 cost questions are still architecture questions. The cheapest answer on paper is wrong if it breaks the access pattern.
| Access pattern | Usually strongest fit | Why |
|---|---|---|
| Object reads and writes over HTTP APIs | S3 | Cheapest common answer when a filesystem is not required |
| Shared POSIX-style file system | EFS or FSx, depending workload | File semantics cost more because the capability is different |
| Low-latency attached block storage | EBS | Block storage is for instance-attached needs, not general object archival |
The exam often rewards the team that avoids using EFS or provisioned SSD storage when ordinary object storage or lower-tier block storage would do the job.
| Pattern | Strongest first fit | Why |
|---|---|---|
| Access is unpredictable and admins do not want to guess the right class | Intelligent-Tiering | Lets AWS move objects between access tiers automatically |
| Access drops sharply after a known number of days | Lifecycle transition to IA or Glacier class | Predictable lifecycle is the classic policy answer |
| Data should expire after a retention window | Lifecycle expiration | Prevents silent storage growth over time |
| External users should pay their own retrieval and transfer cost | Requester Pays, when the business model fits | SAA-C03 may test access-cost allocation, not just storage class |
flowchart LR
S["S3 Standard"] --> I["Standard-IA or Intelligent-Tiering"]
I --> G["Glacier Flexible Retrieval"]
G --> D["Glacier Deep Archive"]
What to notice:
1{
2 "Rules": [
3 {
4 "ID": "move-logs-to-colder-storage",
5 "Status": "Enabled",
6 "Filter": { "Prefix": "logs/" },
7 "Transitions": [
8 { "Days": 30, "StorageClass": "STANDARD_IA" },
9 { "Days": 180, "StorageClass": "GLACIER" }
10 ]
11 }
12 ]
13}
What to notice:
| Requirement | Strongest first fit | Why |
|---|---|---|
| Online transfer from on-premises storage with managed scheduling and verification | DataSync | Better fit than custom copy tooling for recurring transfer jobs |
| Local application needs on-premises interface to AWS-backed storage | Storage Gateway | Hybrid-storage answer rather than raw transfer tool |
| Durable ongoing backup to AWS | AWS Backup or service-native backup pattern | Backup design is separate from primary storage tier |
| Very large one-time migration where network is the bottleneck | Lowest-cost transfer method that fits the time window | The exam may ask whether network copy or another migration path is more economical |
EBS cost questions are often less about one named product and more about not overbuying:
AWS Budgets, Cost Explorer, CUR, cost allocation tags, and multi-account billing are not separate finance trivia. SAA-C03 can use them as the operational answer when the team needs visibility into which buckets, backups, or business units are creating the spend.
| Symptom | Strongest first check | Why |
|---|---|---|
| Old objects pile up in S3 Standard forever | Lifecycle and retention policy | Cost drift often comes from missing automation |
| Team picked EFS for a workload that only needs object storage | Storage type fit | Paying for file-system semantics without needing them is a classic waste pattern |
| Migration scripts are expensive to run and constantly fail | Managed transfer service fit | DataSync or Storage Gateway may be the actual cost and reliability fix |
| Backup bill keeps growing even after primary storage is optimized | Backup retention and archival policy | Primary storage tuning does not automatically fix backup cost |
Continue with 4.2 Compute Solutions to apply the same cost lens to instance families, purchasing models, and elastic runtime choices.