S3 Flashcards
1
Q
What is S3?
A
S3 stands for Simple Storage Service and essentially it’s object storage in the cloud.
2
Q
What kind of storage is S3?
A
Object-based storage
3
Q
How much can you store in S3?
A
Unlimited storage
4
Q
What can you upload to S3?
A
Anything. File, pictures, videos, etc.
5
Q
What can you not us S3 for?
A
Installing an OS, running a DB, etc.
6
Q
Max size of S3 object?
A
5TB
7
Q
What does it mean when S3 is called a universal namespace?
A
The S3 bucket must have a globally unique name as S3 bucket names are global.
8
Q
What is the default bucket URL?
A
It is https:// then the name of the bucket, then .s3, and then . the region that it’s going to be in. A bucket called “test” would have a URL of https://test.s3.us-east-1
9
Q
What response code will you get when successfully uploading a file to S3?
A
HTTP 200 code
10
Q
What does it mean for S3 to be a key-value store?
A
The key-value is the name of the object. If you have an object called “test.jpeg” then the key-value is “test.jpeg”
11
Q
What is a version ID?
A
If versioning is enabled, anytime a new version of an existing object is uploaded, the prior version becomes the old version with a unique version ID that can be restored.
12
Q
What is metadata?
A
It is data about the data.
13
Q
How does S3 ensure durability and availability?
A
S3 data is always spread across multiple devices and facilities
14
Q
S3 is designed for how many 9s of durability?
A
It is built for 11 9’s of durability. 99.999999999
15
Q
True/False. S3 is designed for frequently accessed data.
A
True.
16
Q
Use case examples for S3?
A
websites, content distribution, mobile and gaming applications, big data
17
Q
How can you secure buckets?
A
Server-side encryption, ACLs, and bucket policies
18
Q
What are bucket policies?
A
Actions that can be allowed or denied through a JSON policy that are bucket wide?
19
Q
What are bucket ACLs?
A
Grants permission on a per object bases for users and groups.
20
Q
What is the consistency model of S3?
A
Strong read after write.
21
Q
Can S3 buckets be deployed to a specific region?
A
Yes, but the name space is global.
22
Q
Are buckets private when first created?
A
Yes. They have to be enabled for public access including the objects within them.
23
Q
Can versioning be disabled once it is enabled?
A
No. It can only be suspended.
24
Q
How can you prevent your objects from being accidentally deleted?
A
Enable MFA. Enable versioning.
25
Are versioned objects publicly accessible in a public bucket?
No.
26
Can Lifecycle rules be configured for versioning?
Yes.
27
What are the different S3 storage classes?
S3 Standard, S3 Infrequently Accessed, S3 One Zone-Infrequently Accessed, S3 Intelligent-Tiering
28
What is S3 Standard?
Provides you with high availability and durability, and data is stored redundantly across multiple devices in multiple facilities. Equal to or greater than 3 availability zones. Comes with 99.99% availability and it comes with a 11 nines durability. Designed for frequent access, and it's perfect for frequently accessed data, and it's suitable for most workloads, and this is basically the default storage class. So when people are talking about S3, they're talking about S3 Standard.
29
What are examples of S3 standard?
Place to host video files, or pictures, or any kind of static object. Can be used to host static web sites.
30
What is S3 Standard-Infrequently Accessed (IA)?
Basically just S3 Standard-IA is what it's referred to. So it's designed for infrequently access data. What it does give you though is rapid access. So it's used for data that is accessed less frequently but requires rapid access when you need it. And you basically pay to access the data. So there is a low per-gigabyte storage price and a per-gigabyte retrieval fee. Now, it comes with 99.9% availability and again, 11 nines durability.
31
What are examples of SE Standard-IA?
Use case is where basically you need long-term storage for things like backups, for example, and data stores for disaster recovery files, etc., but you want to minimize your cost, but you still need instant access to that data. That's where you'd use S3 Standard-Infrequently Accessed
32
What is S3 Intelligent-Tiering?
Automatically moves your data to the most cost-effective tier based on how frequently you access each object. It has 99.99% availability, and 11 nines durability. So it's basically using AI and machine learning to figure out how frequently you access your data, and then it will store it on the storage class that's going to save you the most money based off your access patterns.
33
What is S3 One Zone-Infrequently Accessed?
The same is basically the same as S3 Standard-IA. Designed for infrequently access data that gives you rapid access. So it's used for data that is accessed less frequently but requires rapid access when you need it. And you basically pay to access the data. So there is a low per-gigabyte storage price and a per-gigabyte retrieval fee. Now, it comes with 99.9% availability and again, 11 nines durability. It only uses one AZ but stored redundantly within that AZ and is for non-critical data.
34
What are examples of S3 One Zone-Infrequently Accessed?
35
What are examples of S3 Intelligent Teiring?
36
What is Glacier?
Glacier is a way of archiving your data long term, so storing your data for long terms. You pay for each time you access your data. And it's used only for archiving data. So Glacier is really cheap storage, and it's optimized for data that is very infrequently accessed. There are 3 tiers.
37
which is called Glacier Instant Retrieval?
long-term data archiving, but you still have instant retrieval time for your data. So if you've got critical data that you probably might not need to use, maybe only once a year, but you need to access it instantly, then you want Glacier Instant Retrieval.
38
What is Glacier Flexible Retrieval?
This is the ideal storage class for archived data that does not require immediate access but needs the flexibility to retrieve large sets of data at no cost, such as backup or disaster recovery use cases. And it can be minutes in terms of retrieval time or up to 12 hours. So you'd only use this if you can afford to wait for 12 hours to access this data.
39
What is Glacier Deep Archive?
This is the cheapest storage class on AWS, and it's designed for customers that retain datasets for 7-10 years or longer to meet customer needs and regulatory requirements. And the standard retrieval time is around 12 hours but for bulk retrievals, it's around 48 hours.
40
What is the availability and durability of Glacier?
It's the same across all 3 options. So it's 99.99% availability and 11 nines durability.
41
How is S3 priced?
So for S3, the most expensive one is just S3 Standard.
If you want to intelligentlyncontrol your costs and optimize your costs for known or unknown access patterns, then you want S3 Intelligent-Tiering. That's going to move it around between Infrequently Accessed and then just Standard S3.
If you need to access your data and it sits on S3 in Standard-Infrequently Accessed as well as One Zone-Infrequently Accessed, then a retrieval fee is going to apply.
42
How is Glacier priced?
The most expensive one is S3 Glacier. So that's just your standard Glacier. Flexible Retrieval is slightly cheaper,
and then the cheapest is going to be Glacier Deep Archive and again, just remember, your retrieval time for Deep Archive is between 12-48 hours. Glacier Flexible Retrieval can be minutes but it could be up to 12 hours, whereas with normal Glacier, then it's going to be instant access. And a retrieval fee applies for all 3 storage classes.
43
What is S3 Lifecycle Management?
Moving your objects between the different storage tiers, thereby maximizing your cost effectiveness. It's a way of automating, moving your objects to save you money. So if you're not using an object or it hasn't been accessed recently, you can move it from S3 Standard to S3 Infrequently Accessed. So say if it hasn't been accessed in the last 30 days, it would automatically move to S3-IA. And then if you haven't used that object in let's say 90 days after it's been moved to S3-IA, it would then be moved onto Glacier and archived off.
44
Can you use Lifecycle Management with versioning?
You can use Lifecycle Management to move different versions of your objects to different storage tiers as well, thereby saving you even more money.
45
What is S3 Object Lock?
You use it to store objects using a WORM model. And a WORM model just stands for write once, read many. So you can only write once to your objects, but you can read them many, many times. And it can help prevent objects from being deleted or modified for a fixed amount of time or indefinitely.
46
In a scenario, where would you apply the WORM (write once and read many) to Glacier?
In a scenario-based question, they're talking about Glacier, and you're going to want to store it in Glacier Vault Lock.
47
Can Object Lock can be done on individual objects or can it be applied across the bucket as a whole?
Object Lock can be done on individual objects, or it can be applied across the bucket as a whole. And Object Locks come in 2 modes which we covered off. So we have governance mode and compliance mode.
48
Object Locks come in 2 modes which are?
Governance mode and compliance mode.
49
What is Compliance Mode for S3 Object Locks?
Compliance Mode is simply where a protected object version can't be overwritten or deleted by any user whatsoever.
50
What is governance mode for S3 Object Locks?
Prevents most users but you can have some users who are able to either alter an object or delete it.
51
In a scenario, where would you apply WORM in S3?
S3 Object Lock
52
How many types of server-side encryption are there in S3? What are they?
3. SSE, SSE-KMS, SSE-C
53
What is SSE?
SSE is server-side encryption S3. So S3 is where S3 manages the keys. AWS manages all the encryption and decryption for you. You don't need to worry about keys or anything like that. It uses AES 256-bit encryption and this is the most common type of encryption. It's the easiest to use.
54
What is SSE-KMS?
The KMS just stands for Key Management Service. And this is essentially where you work with Amazon on your own key management using the Key Management Service. So you can use KMS keys to encrypt your objects as well.
55
What is SSE-C?
And then we have server-side encryption C, or SSE-C, which is where you provide the keys.
56
How can you enforce server-side encryption?
You can do it using the console, and this is where you select the encryption setting on your S3 bucket. This is the easiest way to do this is just a checkbox in the console. And then we also have it through bucket policy. So you can also enforce encryption using a bucket policy. And this method sometimes comes up in the exam.
57
What is a prefix?
It's simply the folders and sub-folders within your S3 bucket.
58
What is the performance of a prefix?
The more prefixes that you use with your applications using S3, the better performance you're going to get. So you can always achieve a high number of requests. You can do 3,500 puts, copies, posts, and deletes and 5,500 get and head requests per second per prefix. So the more prefixes you have, the faster performance. So you spread your data and your reads across different prefixes.
I gave the example 5,500, so if you're using 2 prefixes, you could then achieve 11,000 requests per second. If you're doing it with 4 prefixes, you could achieve 22,000 requests per second.
59
What are the builtin limits of SSE-KMS?
So uploading and downloading data will count towards the KMS quota. It is region specific, but it's either going to be 5,500,
10,000, or 30,000 requests per second. And currently you can't request a quota increase for KMS.
60
How can you increase upload performance for S3?
You can use multipart uploads to increase your performance when uploading your files to S3.If there's any file that's over 100 meg, then you should be using it. And if you're doing it for any file over 5 gigs, you're forced to use it.
61
How can you increase download performance for S3?
Use S3 byte-range fetches to increase your performance and it's basically doing the opposite. You're splitting up the files into different byte ranges and you're downloading them all at the same time.
62
What is S3 Replication?
It used to be called Cross Region Replication. So, if they haven't updated the questions and you see the mention of cross region replication, then the language has changed. You can replicate objects from one bucket to another. They can either be in the same region or in different regions.
63
Are existing S3 objects replicated when replication is enabled?
No. Objects in an existing bucket are not replicated automatically and delete markers are not replicated automatically by default. However, it can be enabled.
64
How does S3 encrypted a bucket?
S3 doesn’t offer encryption at a bucket level. Encryption occurs at an object level.
65
What are the 3 types of server-side encryption?
Server-Side Encryption With Customer-Provided Keys (SSE-C)
Server-Side Encryption with Amazon S3-Managed Keys (SSE-S3)
Server-Side Encryption With KMS KEYS Stored in AWS Key Management Service (SSE-KMS)
66
What is S3 standard?
Default storage class.
Data is replicated across at least 3 regions
Provides eleven “nines” of durability (1 object loss/10,000,000 objects/10,000 years).
GB/m fee for data stored, fee per GB transferred out charge, price per 1,000 requests.
Should be used for frequently accessed data which is important and non-replaceable.
67
What is S3 standard IA (Infrequent Access)
Roughly half the storage cost of S3 standard.
Higher per GB data retrieval fee.
Minimum duration charge of 30 days (objects can be stored for less time, but the minimum of 30 days still applies for billing.
Minimum capacity charge of 128KB/object.
Should be used for long-lived data that is important but infrequently accessed.
68
What is S3 one-zone IA?
Similar to S3 Standard IA, but cheaper and less resilient due to only using one AZ within the region
Should be used for long-lived data which is non-critical, replaceable if lost, and infrequently accessed
69
What is S3 Glacier Instant?
Similar to S3 Standard IA, but cheaper storage, more expensive retrieval, and longer minimum storage duration.
Should be used for long-lived data that is only accessed roughly once per quarter and needs millisecond retrieval time.
70
What is S3 Glacier Flexible?
Very cheap, but requires minutes or hours to access data.
Objects cannot be made publicly accessible, any access of data (beyond object metadata) requires a retrieval process.
Should be used for archival data where frequent or realtime access isn’t needed (e.g. yearly).
71
What is S3 Glacier Deep Archive?
Similar to S3 Glacier Flexible, but with longer retrieval times of roughly 12-48 hours.
Should be used for archival data that rarely if ever needs to be accessed.
72
What is S3 Intelligent Tiering?
Monitors and automatically moves any objects between different tiers based on their usage.
Has a monitoring and automation cost per 1,000 objects.
Should be used for long-lived data with changing or unknown usage patterns.
73
Go back and add rest of S3 cards.
Do it.
