10 Storage Services · AWS Cloud Practitioner

In this chapter you will learn the difference between object, block, and file storage; how Amazon S3 works, including buckets, objects, durability, storage classes, lifecycle rules, versioning, replication, static website hosting, Block Public Access, Transfer Acceleration, and S3 Express One Zone; how Amazon EBS volumes, snapshots, and instance store support EC2; when to use Amazon EFS and the Amazon FSx family of file systems; and how AWS Storage Gateway, AWS Backup, and AWS Elastic Disaster Recovery support hybrid storage, backup, and recovery. This chapter maps to Domain 3, Cloud Technology and Services, Task Statement 3.6 (identify AWS storage services).

10.1 Three kinds of storage

Before looking at individual services, it helps to understand the three basic ways computers store data. Almost every storage question on the exam becomes easy once you know which kind of storage the scenario needs. Block storage splits data into fixed-size blocks and presents them to an operating system as a raw disk. The OS formats it with a file system and can change any single block in place. It is fast and low-latency, which makes it ideal for operating systems, databases, and applications that update small pieces of data frequently. On AWS: Amazon EBS and instance store.

File storage organizes data in a hierarchy of folders and files and shares it over a network using protocols such as NFS (Linux) or SMB (Windows). Many servers can read and write the same files at the same time. On AWS: Amazon EFS and Amazon FSx.

Object storage stores each piece of data as a whole object with a unique key and metadata, in a flat namespace, accessed over HTTPS APIs. You do not edit part of an object; you replace the whole thing. It scales almost without limit and is inexpensive. On AWS: Amazon S3.

Figure 10.1: Block, file, and object storage compared. Block storage looks like a local disk to one server, file storage is a shared folder tree for many

servers, and object storage is a flat collection of objects reached over HTTPS.

Exam tip

“Boot volume” or “database disk for an EC2 instance” means block (EBS). “Shared file system for many Linux instances” means EFS; for Windows, FSx for Windows File Server. “Store unlimited images, backups, or logs” or “static website” means object (S3).

10.2 Amazon S3 fundamentals

Amazon Simple Storage Service (S3) is AWS’s object storage service and one of the oldest and most widely used services in AWS. It stores data for websites, mobile apps, backups, data lakes, analytics, machine learning, and much more.

Buckets and objects

Data in S3 lives in buckets. A bucket is a container you create in a specific Region, and its name must be unique across all AWS accounts in the partition (so my-photos is almost certainly taken). Inside a bucket you store objects. Each object has:

A key, its full name, such as 2026/reports/sales.pdf . The slashes make keys look like folders in the console, but S3’s namespace is actually flat.

The data itself, from zero bytes up to multiple terabytes. Large files are uploaded in parts with multipart upload, which is recommended for objects over about 100 MB.

Metadata, such as content type and custom tags.

Optionally, a version ID if versioning is enabled.

There is no limit on the number of objects in a bucket or the total amount stored. You pay for what you store, the requests you make, and data transferred out.

Durability and availability

S3 Standard is designed for 99.999999999% (11 nines) durability of objects over a year. In practical terms, if you store ten million objects, you could expect to lose a single object roughly once every ten thousand years. S3 achieves this by storing data redundantly across multiple devices in at least three Availability Zones in the Region (except the One Zone classes). Durability is about not losing data; availability is about being able to reach it right now. S3 Standard is designed for 99.99% availability.

S3 also provides strong read-after-write consistency: as soon as a write succeeds, any read returns the new data.

Security defaults

New S3 buckets are private by default. S3 Block Public Access is turned on for new buckets, so even if someone writes a bucket policy or ACL that would grant public access, Block Public Access overrides it. It can be set at the bucket level and, more powerfully, at the account level to stop anyone in the account from making data public. New buckets also have access control lists (ACLs) disabled by default, so access is managed through policies, and all new objects are encrypted at rest automatically with S3-managed keys (SSE-S3). You can choose AWS KMS keys (SSE-KMS) instead for extra control and auditing.

Access is controlled through IAM policies (what a user or role can do) and bucket policies (who can access a specific bucket). You can also generate presigned URLs that grant time-limited access to a single object without making it public.

Common trap

Many real-world data leaks came from S3 buckets that were made public by mistake. If a question asks how to prevent any bucket in an account from ever being made public, the answer is S3 Block Public Access at the account level, not encryption. Encryption protects data at rest; it does not stop a public bucket from serving files to anyone who asks. Console S3 > Buckets > Create bucket Region: Canada

Create bucket
General configuration
AWS Region: Canada (Central) ca-central-1
Bucket type General purpose Directory (Express One Zone)
Bucket name yourname-lab-2026-0417
Must be globally unique; lowercase, numbers, hyphens
Object Ownership
ACLs disabled (recommended) ACLs enabled
Block Public Access settings for this bucket
Block all public access
Block public access granted through ACLs
Block public access granted through bucket policies
(and two more options)
Bucket Versioning
Disable Enable
Default encryption
SSE-S3 (Amazon S3 managed keys) SSE-KMS (AWS KMS keys)
Cancel Create bucket

Figure 10.2: The S3 Create bucket screen. You choose a Region and a globally unique name; Object Ownership (ACLs disabled), Block Public

Access, and default encryption are already set to secure defaults. (Illustration, not an actual AWS screenshot.)

Hands-on lab

Hands-on mini lab: create a bucket and upload an object

1. In the console, open the S3 service and choose Create bucket.
2. For Bucket type, keep General purpose.
3. Enter a globally unique name, such as yourname-lab-2026-0417 , and confirm the Region shown.
4. Leave Object Ownership set to ACLs disabled.
5. Leave Block all public access checked.
6. Under Bucket Versioning, choose Enable.
7. Leave Default encryption as SSE-S3 and choose Create bucket.
8. Open the bucket, choose Upload, add a small text file, and choose Upload.
9. Upload a changed file with the same name, then turn on Show versions to see both versions.
10. To clean up, empty the bucket (deleting all versions) and then delete the bucket.

10.3 S3 storage classes

Not all data is accessed equally. Last week’s photos are viewed constantly; last year’s tax records are rarely touched but must be kept. S3 offers storage classes that trade lower storage cost for higher retrieval cost, slower retrieval, or reduced redundancy. You can set the class per object.

Storage class Access pattern Notes

S3 Standard Frequent Default; 3+ AZs; no retrieval fee

S3 Intelligent-Tiering Unknown or changing Moves data between tiersautomatically S3 Express One Zone Very frequent, latency-critical Single-digit ms; one AZ; directory buckets S3 Standard-IA Infrequent, fast when needed Retrieval fee; 30-day minimum

S3 One Zone-IA Infrequent, re-creatable One AZ only; cheaper than Standard-IA
S3 Glacier Instant Retrieval Rare (about quarterly), msaccess 90-day minimum
S3 Glacier Flexible Retrieval Rare; minutes to hours OK 90-day minimum
S3 Glacier Deep Archive Very rare; hours OK Lowest cost; 180-day minimum
A few details are worth understanding rather than memorizing:

Infrequent Access (IA) classes are cheaper to store but charge a per-GB retrieval fee and have a minimum storage duration (you pay for at least 30 days even if you delete sooner). They are great for backups and older data that still needs instant access when it is requested.

One Zone classes keep data in a single AZ. They are cheaper, but if that AZ is destroyed, the data is lost. Use them only for data you can re-create, such as thumbnails or secondary backup copies.

S3 Intelligent-Tiering monitors each object’s access and automatically moves it between frequent, infrequent, and archive-instant tiers. There is a small monthly monitoring fee per object but no retrieval fees. It is the right answer when access patterns are unknown or unpredictable.

The Glacier classes are for archives. Instant Retrieval returns data in milliseconds. Flexible Retrieval returns data in minutes (expedited) to hours (standard or bulk). Deep Archive is the cheapest storage in AWS, with retrieval typically within 12 hours (or up to about 48 hours for bulk), and suits data kept for years for compliance. S3 Express One Zone is the high-performance end: it stores data in a single AZ that you choose, in a special directory bucket, to deliver consistent single-digit-millisecond latency for request-heavy workloads such as machine learning training and analytics. Placing compute in the same AZ gives the best performance.

The S3 storage class ladder
Retrieval cost and wait time increase
Storage price per GB increases
S3 Express One Zone single-digit ms, one AZ, directory bucket
S3 Standard frequent access, ms, no retrieval fee
S3 Intelligent-Tiering unknown patterns: moves tiers for you
S3 Standard-IA infrequent, ms, retrieval fee, 30-day min
S3 One Zone-IA infrequent, re-creatable, one AZ
Glacier Instant Retrieval archive, ms access, 90-day min
Glacier Flexible Retrieval archive, minutes to hours, 90-day min
Glacier Deep Archive lowest cost, about 12-48 hours, 180-day min

Read often: stay near the top. Read rarely and can wait: move down the ladder.

All classes except One Zone and Express One Zone store data across at least three AZs.

Figure 10.3: The S3 storage class ladder. Moving down the ladder lowers the monthly storage price but raises the cost and time to retrieve data, so match the class to how often the data is read.

Exam tip

Look for the access-pattern keyword. “Frequently accessed” means Standard. “Unknown or changing access patterns” means Intelligent-Tiering. “Infrequently accessed but needs immediate access” means Standard-IA (or Glacier Instant Retrieval if accessed only a few times a year). “Can be re-created” plus “infrequent” means One Zone-IA. “Archive, retrieval within hours, lowest cost” means Glacier Deep Archive.

10.4 Managing data in S3

Lifecycle configuration

You do not have to move objects between classes by hand. An S3 Lifecycle configuration is a set of rules applied to a bucket (or to objects with a certain prefix or tag) with two kinds of actions:

Transition actions move objects to a cheaper class after a set number of days.

Expiration actions delete objects (or old versions) after a set number of days.

For example, a company storing application logs might keep them in S3 Standard for 30 days while engineers actively investigate issues, transition them to Standard-IA, move them to Glacier Flexible Retrieval after 90 days, then to Deep Archive after a year, and finally delete them after seven years when the retention requirement ends. Lifecycle rule for application logs (prefix: logs/)

Bar heights are illustrative, not actual prices.

Figure 10.4: An example lifecycle policy for log files. Objects move automatically to cheaper storage classes as they age and are deleted when the retention period ends.

Versioning

S3 Versioning keeps every version of an object. When you overwrite an object, the old version is preserved; when you delete one, S3 adds a delete marker and the earlier versions remain recoverable. Versioning protects against accidental overwrites and deletions and is a building block for replication. Once enabled, versioning can be suspended but not fully turned off. Remember that every version is stored and billed, so pair versioning with lifecycle rules that expire old versions. For stronger protection, S3 Object Lock provides write-once-read-many (WORM) storage that prevents objects from being deleted or changed for a retention period, which helps meet some regulatory requirements.

Replication

S3 Replication automatically copies new objects from one bucket to another. Cross-Region Replication (CRR) copies to a bucket in a different Region, for disaster recovery, compliance requirements to keep copies far apart, or lower latency for users in another geography. Same-Region Replication (SRR) copies within a Region, for example to aggregate logs or keep a copy in a separate account. Both require versioning on the source and destination buckets.

Static website hosting

S3 can host a static website: HTML, CSS, JavaScript, images, and downloadable files that do not need server-side code. You enable static website hosting on the bucket and specify an index document (such as index.html ). In production, the recommended pattern is to keep the bucket private and put Amazon CloudFront in front of it, so content is served over HTTPS from edge locations and only CloudFront can read the bucket.

S3 Transfer Acceleration

S3 Transfer Acceleration speeds up uploads to a bucket from users who are far away. Uploads go to the nearest CloudFront edge location and then travel over the AWS global network to the bucket’s Region. It is useful when customers around the world upload large files to a central bucket. You pay an extra per-GB fee only when acceleration actually improves the transfer.

Real-world example

Real-world scenario: A video production company in Toronto receives large raw footage uploads from film crews in Australia, South Africa, and Brazil. Uploads directly to a bucket in the Canada (Central) Region are slow and often fail. Turning on S3 Transfer Acceleration and using multipart upload lets crews upload through nearby edge locations, cutting transfer times significantly without changing where the data is stored.

10.5 Amazon EBS

Amazon Elastic Block Store (EBS) provides persistent block storage volumes for EC2 instances. An EBS volume behaves like a hard drive attached to your instance, but it is actually network-attached storage, which means it persists independently of the instance’s life: you can stop an instance, detach a volume, and attach it to another instance.

AZ-scoped

An EBS volume lives in one Availability Zone and can attach only to instances in that same AZ. Within the AZ, EBS automatically replicates the volume to protect against component failure. To move data to another AZ or Region, you use a snapshot. Normally a volume attaches to one instance at a time; certain Provisioned IOPS volumes support Multi-Attach to several instances in the same AZ for specialized clustered applications.

Volume types
EBS volumes come in SSD-backed and HDD-backed types:
Type Backed by Best for
gp3 (General Purpose SSD) SSD Most workloads, boot volumes; default
choice
io2 Block Express/io1 (Provisioned IOPS
SSD)
SSD Critical, I/O-intensive databases
st1 (Throughput Optimized HDD) HDD Big data, log processing, sequential reads
sc1 (Cold HDD) HDD Infrequently accessed data, lowest cost

gp3 is the usual default. It provides a solid baseline of performance and lets you add IOPS and throughput independently of size. io2 volumes deliver the highest, most consistent performance and durability for demanding databases. The HDD types are cheaper per GB but cannot be used as boot volumes. (The older gp2 type still exists.) Snapshots

An EBS snapshot is a point-in-time backup of a volume, stored by AWS in S3-backed storage that you do not manage directly. Snapshots are incremental: after the first full snapshot, each new one saves only the blocks that changed, which keeps costs down. You can use a snapshot to restore a volume, create a volume in a different AZ, or copy it to another Region for disaster recovery. Amazon Data Lifecycle Manager or AWS Backup can create and delete snapshots on a schedule.

EBS volumes and snapshots can be encrypted with AWS KMS keys. Encryption covers data at rest, data moving between the instance and the volume, and snapshots created from the volume.

Instance store

Instance store volumes are disks physically attached to the host computer that runs your instance, available on certain instance types. They offer very high I/O performance at no extra charge, but the data is temporary: it is lost when the instance stops, hibernates, or terminates, or if the underlying drive fails. (It survives a simple reboot.) Use instance store for buffers, caches, scratch data, and data replicated elsewhere, never as the only copy of important data.

10.6 File storage: Amazon EFS and Amazon FSx

Amazon EFS

Amazon Elastic File System (EFS) is a fully managed, elastic NFS file system for Linux workloads. You create a file system and mount it on many EC2 instances, containers, or Lambda functions at once, across multiple AZs. It grows and shrinks automatically as you add and remove files, so you never provision capacity, and you pay for the storage you actually use.

EFS is a natural fit when several servers need the same files: content management systems, shared web content, home directories, development tools, and machine learning data. To cut costs, EFS offers storage classes (Standard, Infrequent Access, and Archive) plus lifecycle management that moves files that have not been read recently to the cheaper classes. A One Zone option stores data in a single AZ for lower cost.

EBS: one AZ, normally one instance. EFS: many instances, many AZs.

Figure 10.5: EBS versus EFS. An EBS volume attaches to an instance in the same Availability Zone, while one EFS file system is mounted by many instances across multiple Availability Zones.

Amazon FSx

Amazon FSx provides fully managed file systems built on popular commercial and open-source file systems. Choose by name:

FSx for Windows File Server: native Windows file shares over SMB, with Microsoft Active Directory integration, for Windows applications, home directories, and lift-and-shift of Windows file servers.

FSx for Lustre: a high-performance parallel file system for HPC, machine learning, and media processing. It can link to an S3 bucket, presenting S3 objects as files and writing results back.

FSx for NetApp ONTAP: NetApp’s ONTAP file system, supporting NFS, SMB, and iSCSI, for migrating existing NetApp workloads with familiar features.

FSx for OpenZFS: the OpenZFS file system over NFS, for migrating Linux file servers that use ZFS.

Exam tip

“Windows,” “SMB,” or “Active Directory” file shares mean FSx for Windows File Server. “High-performance computing” or “Lustre” means FSx for Lustre. “Linux,” “NFS,” “shared across many instances,” and “elastic” with no other brand mentioned mean EFS.

10.7 Hybrid storage with AWS Storage Gateway

Many organizations are not ready to move every application to AWS but want to use cloud storage from their data centers, for example to replace aging tape libraries or to get off-site backup. AWS Storage Gateway is a hybrid cloud storage service that gives on-premises applications access to virtually unlimited cloud storage. You run the gateway as a virtual machine (or hardware appliance) in your data center, and it presents standard storage protocols to local applications while storing data in AWS. It caches frequently used data locally for low-latency access. There are several gateway types:

Amazon S3 File Gateway: presents an NFS or SMB file share; files are stored as objects in S3, where you can apply lifecycle rules and analytics.

Amazon FSx File Gateway: provides low-latency on-premises access to FSx for Windows File Server shares (check current availability, as AWS has limited this type for new customers).

Volume Gateway: presents iSCSI block volumes. In cached mode, the primary data is in AWS with frequently accessed data cached locally; in stored mode, the full dataset is local and backed up asynchronously to AWS as EBS snapshots.

Tape Gateway: presents a virtual tape library to existing backup software, storing virtual tapes in S3 and archiving them to the Glacier classes. It replaces physical tapes without changing backup processes.

Nfs / smb

iSCSI
virtual tape (VTL)
Applications use standard

protocols; no code changes.

Hybrid storage: keep apps on premises, put capacity and durability in AWS.

Figure 10.6: AWS Storage Gateway in a hybrid architecture. On-premises servers and backup software talk to a local gateway using file, block,

or tape protocols, and the gateway stores data in Amazon S3, S3 Glacier, or Amazon FSx over the network.

10.8 Backup and disaster recovery

AWS Backup

AWS Backup is a fully managed service that centralizes and automates backups across many AWS services, including EC2 and EBS, RDS and Aurora, DynamoDB, EFS, FSx, S3, and Storage Gateway volumes, as well as some on-premises workloads. Instead of configuring backups separately in each service, you create a backup plan that defines frequency, retention, and lifecycle (for example, “daily backups kept for 35 days, monthly backups moved to cold storage and kept for seven years”) and assign resources to it, often by tag.

Backups are stored in backup vaults. You can copy backups to other Regions and other accounts for resilience against regional events or a compromised account. AWS Backup Vault Lock enforces write-once-read-many protection so that backups cannot be deleted before their retention ends, even by administrators, which helps against ransomware. With AWS Organizations, backup policies can be applied across all accounts, and AWS Backup Audit Manager reports on compliance.

AWS Elastic Disaster Recovery

AWS Elastic Disaster Recovery (AWS DRS) minimizes downtime and data loss for servers, whether they run on premises, in another cloud, or in another AWS Region. It installs a replication agent on each source server and continuously replicates its disks at the block level into a low-cost staging area in your AWS account. Because you are not paying for full-size duplicate servers that sit idle, the cost stays low. When disaster strikes, or during a drill, you launch recovery instances in minutes from the latest state or a chosen point in time. AWS DRS supports a recovery point objective (RPO) measured in seconds and a recovery time objective (RTO) measured in minutes. After the event, you can fail back to the original site.

Need Service

Central, policy-based backupsacross services AWS Backup

Fast recovery of whole servers with seconds of dataloss AWS Elastic Disaster Recovery Point-in-time copy of one EBS volume EBS snapshot

Keep previous versions of S3 objects S3 Versioning

Copy objects to another Region automatically S3 Cross-Region Replication

Common trap

Replication is not the same as backup. If a file is deleted or corrupted in the source, replication may copy that change to the destination. Backups (AWS Backup, snapshots, S3 Versioning) keep earlier points in time that you can restore from.

Chapter 10 summary

Block storage (EBS, instance store) acts as a disk for one instance; file storage (EFS, FSx) is a shared folder tree; object storage (S3) holds whole objects in a flat namespace over HTTPS.

S3 stores objects in Regional buckets with globally unique names, is designed for 11 nines of durability, and is private by default, with Block Public Access and default encryption.

Storage classes range from Standard and Express One Zone to Intelligent-Tiering, Standard-IA, One Zone-IA, and the three Glacier classes; match the class to the access pattern. Lifecycle rules transition and expire objects; versioning protects against overwrites and deletes; CRR/SRR copy objects to other buckets; Transfer Acceleration speeds long-distance uploads; S3 can host static websites. EBS volumes are persistent, AZ-scoped block storage; gp3 is the default, io2 suits demanding databases, and snapshots are incremental and can be copied across Regions.

Instance store is fast but ephemeral; never keep the only copy of important data there.

EFS is elastic NFS for Linux shared across many instances and AZs; FSx offers Windows File Server, Lustre, NetApp ONTAP, and OpenZFS.

Storage Gateway connects on-premises applications to AWS storage through file, volume, and tape gateways. AWS Backup centralizes backup policies; AWS Elastic Disaster Recovery continuously replicates servers for fast recovery.

Key terms
Term Meaning
Object storage Whole objects with keysand metadata, accessed by API
Bucket Regional S3 container with a globally unique name
Durability Likelihood datais not lost (S3: 11 nines)
Storage class S3 tier balancing storageand retrieval cost
Lifecycle rule Automatic transition or expiration of objects
Versioning Keepsall versions of S3 objects
Block Public Access Settings that overrideany public access to S3
Amazon EBS Persistent block volumes for EC2, in one AZ
Snapshot Incremental point-in-time backup ofan EBS volume
Instance store Temporary block storage on the host
Amazon EFS Elastic, shared NFS file system for Linux
Amazon FSx Managed Windows, Lustre, ONTAP, OpenZFS file systems
Storage Gateway Hybrid access from on premises to AWS storage
AWS Backup Central, policy-based backup service
Elastic Disaster Recovery Continuous server replication for fast recovery
Practice questions

1. A company needs to store millions of product images that will be served to a website, with virtually unlimited capacity and high durability. Which service should it use?

A. Amazon EBS B. Amazon S3
C. Amazon EC2 instance store D. Amazon FSx for Lustre

2. What durability is Amazon S3 Standard designed to provide for objects over a given year?

A. 99.9% B. 99.99%
C. 99.999999999% D. 100%

3. A company stores data whose access patterns are unknown and change over time. It wants to optimize storage costs automatically without retrieval fees. Which S3 storage class should it use?

A. S3 Standard-IA B. S3 Intelligent-Tiering
C. S3 Glacier Deep Archive D. S3 One Zone-IA

4. A company must keep financial records for seven years. The records are almost never accessed, and retrieval within 48 hours is acceptable. Which storage class is the most cost-effective?

A. S3 Standard B. S3 Glacier Instant Retrieval
C. S3 Glacier Deep Archive D. S3 Express One Zone

5. A company stores thumbnail images that can easily be regenerated from originals. The thumbnails are accessed infrequently. Which storage class offers the lowest cost while still providing millisecond access?

A. S3 One Zone-IA B. S3 Standard
C. S3 Glacier Flexible Retrieval D. S3 Intelligent-Tiering

6. A company wants objects in a bucket to move automatically to S3 Standard-IA after 30 days and be deleted after one year. What should it configure?

A. S3 Versioning B. S3 Lifecycle rules
C. S3 Cross-Region Replication D. S3 Transfer Acceleration

7. Which S3 feature protects against accidental overwrites and deletions by retaining previous copies of objects?

A. S3 Block Public Access B. S3 Versioning
C. S3 Transfer Acceleration D. S3 static website hosting

8. A security team must ensure that no S3 bucket in an AWS account can be made publicly accessible, even if a user writes a public bucket policy. What should they enable?

A. Default encryption with SSE-S3 B. S3 Block Public Access at the account level
C. S3 Versioning D. MFA on the root user

9. Users around the world upload large files to a single S3 bucket in one Region and experience slow uploads. Which feature will help most?

A. S3 Cross-Region Replication B. S3 Transfer Acceleration
C. S3 One Zone-IA D. S3 Lifecycle rules

10. Which statement about Amazon EBS volumes is correct?

A. A volume can be attached to instances in any Availability

Zone in the Region

B. A volume exists in a single Availability Zone and attaches

to instances in that AZ

C. Volumes are deleted automatically when an instance is

stopped

D. Volumes are object storage accessed over HTTPS

11. A company wants to back up an EBS volume and restore it in another AWS Region for disaster recovery. What should it do?

A. Enable S3 Versioning on the volume B. Create an EBS snapshot and copy it to the other Region
C. Move the volume to instance store D. Enable Multi-Attach

12. An application on EC2 needs temporary scratch space with the highest possible I/O performance. The data can be lost without harm. Which storage is the best fit?

A. Amazon S3 Standard B. Instance store
C. Amazon EFS D. S3 Glacier Instant Retrieval

13. Dozens of Linux EC2 instances across multiple Availability Zones need to read and write the same set of files at the same time. Which service should be used?

A. Amazon EBS gp3 B. Amazon EFS
C. Instance store D. AWS Storage Gateway Tape Gateway

14. A company is migrating Windows-based applications that rely on SMB file shares integrated with Microsoft Active Directory. Which service fits best?

A. Amazon EFS B. Amazon FSx for Windows File Server
C. Amazon FSx for Lustre D. Amazon S3

15. A company wants to replace its physical tape library with cloud storage without changing its existing backup software. Which service should it use?

A. AWS Storage Gateway Tape Gateway B. Amazon EBS snapshots
C. Amazon EFS Infrequent Access D. AWS Elastic Disaster Recovery

16. A company wants a single place to define backup schedules and retention policies for EC2, RDS, DynamoDB, and EFS across multiple accounts. Which service should it use?

A. AWS Backup B. Amazon S3 Lifecycle
C. AWS Storage Gateway D. Amazon Data Lifecycle Manager

17. A company wants to recover its on-premises application servers in AWS within minutes, losing only seconds of data, while keeping costs low until a disaster occurs. Which service fits best?

A. AWS Backup B. AWS Elastic Disaster Recovery
C. S3 Same-Region Replication D. Amazon FSx for NetApp ONTAP

18. Which use cases are good fits for Amazon S3? (Choose two.)

A. Hosting a static website B. The boot volume of an EC2 instance
C. Storing backups and data lake files D. A shared NFS file system mounted by Linux servers
E. Temporary scratch space attached to an instance

Answers and explanations

1. B. S3 offers virtually unlimited, highly durable object storage that suits images served by websites. EBS and instance store are block storage for a single instance, and FSx for Lustre is a high-performance file system.
2. C. S3 Standard is designed for 11 nines (99.999999999%) of durability. 99.99% is its designed availability, a different measure.
3. B. Intelligent-Tiering moves objects between tiers automatically based on access, with no retrieval fees. Standard-IA and One Zone-IA charge retrieval fees and assume a known infrequent pattern.
4. C. Glacier Deep Archive is the lowest-cost class for long-term archives with retrieval in hours. Standard and Instant Retrieval cost more to store, and Express One Zone is for high performance.
5. A. One Zone-IA is cheaper than Standard-IA, offers millisecond access, and suits re-creatable data. Glacier Flexible Retrieval is not millisecond access.
6. B. Lifecycle rules provide transition and expiration actions. Versioning, replication, and Transfer Acceleration do not move or delete objects on a schedule.
7. B. Versioning keeps prior versions and uses delete markers so objects can be recovered. Block Public Access controls exposure, not recovery.
8. B. Account-level Block Public Access overrides any policy or ACL that would make buckets public. Encryption and root MFA do not prevent public access.
9. B. Transfer Acceleration routes uploads through nearby edge locations and the AWS network. CRR copies data after it arrives and does not speed uploads.
10. B. EBS volumes are AZ-scoped and attach to instances in the same AZ. Volumes persist when an instance stops, and they are block storage, not object storage.
11. B. EBS snapshots can be copied to another Region and used to create new volumes there. Versioning is an S3 feature, and Multi-Attach is within one AZ.
12. B. Instance store provides very fast, temporary storage physically attached to the host. S3 and EFS are network storage with higher latency.
13. B. EFS is a shared, elastic NFS file system that many Linux instances across AZs can mount concurrently. EBS volumes are tied to one AZ and normally one instance.
14. B. FSx for Windows File Server provides native SMB shares with Active Directory integration. EFS is NFS for Linux, and FSx for Lustre targets HPC.
15. A. Tape Gateway presents a virtual tape library to existing backup software and stores tapes in S3 and Glacier. Elastic Disaster Recovery replicates servers, not tapes.
16. A. AWS Backup centralizes backup plans across many services and accounts. Data Lifecycle Manager handles only EBS snapshots and EBS-backed AMIs, and S3 Lifecycle only manages S3 objects.
17. B. Elastic Disaster Recovery continuously replicates servers to a low-cost staging area and recovers in minutes with an RPO of seconds. AWS Backup restores from periodic backups with longer RPO and RTO.
18. A and C. S3 hosts static websites and is widely used for backups and data lakes. Boot volumes and scratch space need block storage, and shared NFS needs EFS.