Chapter 3 Hello, Storage Flashcards

1
Q

Enumerate data value types from lowest to highest relevance

A
  1. Transient
  2. Reproducible
  3. Authoritative
  4. Critical/Regulated
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What can you do on an existing EBS volume, and would these changes require service interruption?

A
  1. increase size
  2. modify IOPS capacity
  3. change volume type

No service interruptions are required

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Why are EBS snapshots cost effective?

A

They implement incremental backup strategies. So, for example, if 5GB of a 100GB volume had changed, the backup is only 5GB large.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Amazon EBS Optimization

A

This is an EBS features which allows an instance to prevent network contention for an EBS volume. It will make sure EBS related traffic is kept separate from general network traffic.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What is “initialization” in regard to snapshots?

A

EBS volumes restored from snapshots are slower than a new EBS volume at the point at which a block is accessed for the first time. If you want to avoid this performance penalty, you can ‘initialize’ the volume by accessing each block before putting it into production

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What data access behavior is sub optimal for HDD?

A

small, random I/O

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Strategies for degraded HDD performance?

A
  1. Configure read-ahead for ready heavy workloads
  2. use RAID 0 configuration
  3. track performance via Amazon CloudWatch
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What to do if EBS is root volume for instance but instance is inacessible?

A

Make new EC2 instance, detach EBS volume from failing instance and attach to new instance

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Describe bucket limitations

A
  1. They are flat (no nested buckets)
  2. only 100 buckets per accounts
  3. must be empty to delete it
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

What happens when you delete an object in a versioned bucket?

A

A delete ‘marker’ is applies on the object, so it is inaccessible. But administrators can still access the data.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What is reason for versioning in S3?

A
  1. protect from accidental deletion
  2. recover earlier version
  3. retrieve deleted objects
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Can you revert a bucket to an unversioned state?

A

No, but you can copy all the versioned objects into a new unversioned bucket and use that bucket going forward

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Benefits of tagging objects

A
  1. Can create fine grained access (i.e. permission for certain tags)
  2. Fine grained control in lifecycle management
  3. Filters in S3 Analytics
  4. Customize CloudWatch Metrics based on tag filters
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

What is a vault lock?

A

A set of policies used for compliance on S3 Glacier Vault. An example policy is ‘write once read many’. A a vault lock is permanent.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

What is the base unit of storage in S3 Glacier?

A

An archive

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Can you assign a key name to S3 Glacier Archives?

A

No, you simply receive a sequence of characters as an identifier when you upload

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

What is minimum # of AZs used by all S3 storage classes (except OneZone_IA)

A

3

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

What data consistency model does S3 follow?

A

Eventual consistency, but it uses read after write consistency for PUT operations of new objects.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

What is envelope encryption?

A
  1. A data key is generated
  2. Data is encrypted with data key.
  3. Data key itself is encrypted with existing encryption key.
  4. Encrypted data and data key stored by AWS storage service.

*The key encrypting key are stored and managed separately from the data and data keys

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

Options for S3 encryption?

A
  1. SSE-S3
  2. SSE-C (customer managed keys)
  3. SSE-KMS
21
Q

Default access policy for a bucket?

A

private

22
Q

Difference b/t bucket policies and IAM policies?

A

Instead of attaching policies to the users, groups, or roles, bucket policies are attached to a specific resource, such as an Amazon S3 bucket.

23
Q

T/F: you can grant access to other accounts and your account using resource based ACLs

A

False, you can only grant access to other accounts using resource based ACLs, not users in your own account.

24
Q

Defense in depth strategies for S3

A
  1. Data must be encrypted at rest and in transit
  2. Data must be accessible only by a limited set of public IP addresses
  3. Data must not be publicly accessible directly from an S3 URL
  4. Domain name required to consume content
  5. Apply bucket policies for users w/ specific permissions
25
Q

Another name for query string authentication of S3

A

presigned URL

26
Q

What does Cross Region Replication (CRR) apply to and what are some of its characteristics?

A

Replication of objects in distinct regions

  1. S3 must be granted permission to replicate
  2. Versioning must be enabled
  3. Source and destination must be in different regions
27
Q

What to enable with S3 for GET intensive workloads?

A

CloudFront

28
Q

What strategy to implement for heavy worklaods in S3 >= 1,000 request per second?

A

Implement a short (3-4 character) random hash into the key prefix for an S3 object. This ensures that objects will not reside in close proximity within same infrastructure partition.

29
Q

What to enable to transfer large objects across large distances? How does this solution work?

A

S3 Transfer Acceleration. Works by transferring data to nearest CloudFront Edge location. Then data can avoid public internet when transferring to the target region (lower latency).

30
Q

What character to avoid in S3 Buckets?

A

periods

31
Q

How to upload very large files to S3?

A

multipart uploads (recommended for >= 100MB)

  1. break apart file
  2. parallelize upload
  3. submit manifest file
32
Q

How to download very large file from S3?

A

Range GET. This is basically the opposite of multipart uploads.

33
Q

Difference b/t EBS and S3 billing.

A

In S3, you pay for what is used; in EBS you pay for what is allocated.

34
Q

Lifecycle Management option for S3

A
  1. Transition Actions (i.e. move to STANDARD_IA after 30 days)
  2. Expiration Actions (i.e. delete after 30 days)
35
Q

A key limitation of EFS?

A

Not supported on Windows instances

36
Q

How to sync files with EFS?

A

AWS DataSync

37
Q

Performance options within EFS?

A
  1. General Purpose (low latency)

2. Max I/O (higher latency, more scalable)

38
Q

File gateway is like…

A

an NFS mount on Amazon S3, allowing you to access your data directly in Amazon S3 from on premises as a file share.

39
Q

Volume Gateway provides…

A

cloud-based storage volumes you can mount as iSCSI devices from on prem application servers.
(cached mode and stored volume mode configurations available)

  1. Cached mode: frequently accessed data cached in gateway
  2. stored volume mode: volumes asynchronously backed up to cloud
40
Q

Snowball is for ___ scale migrations, and Snowmobile is for __ scale migrations.

A

Snowball: petabyte (< 10 PB)
SnowMobile: exabyte ( > 10 PB)

41
Q

Difference b/t AWS managed VPN and AWS VPN Cloudhub

A

Both offer tunneling into VPC (exposes VPN endpoints via virtual private gateway), but Cloudhub is designed for when you have more than one remote network (i.e. multiple branch offices)

42
Q

S3 durability and availability

A

Durability: 11 nines
Availability: 4 nines

43
Q

One way to protect against accidental object deletion that is not versioning?

A

MFA delete

44
Q

Glacier size limit?

A

40TB

45
Q

Storage limits

A
  1. EC2 instance store: 48TB
  2. EBS: 16TB
  3. S3: limitless
  4. Glacier: limitless
46
Q

When restoring an EBS volume from a snapshot, how long will it take before data is available?

A

The data will be available immediately

47
Q

Steps to create static website in S3

A
  1. Create bucket with website hostname
  2. Upload static content and make it public
  3. Enable website hosting on bucket
  4. indicate index and error page details
48
Q

When you delete an EC2 instance, is EBS volume deleted?

A

If the EBS volume is root, then yes; otherwise no. Can be manually set via DeleteOnTermination attribute.