[AWS Solution Architect] S3

S3 (Simple Storage Service)

• it provides you with unlimited storage
• The S3 console provides an interface for you to upload and access your data.
• S3 Object : contain data like files.
  • You can store data from 0 bytes to 5 Terabytes in size.
• S3 Bucket : hold objects like folders.
  • Bucket names must be unique.

Object-based storage service

1

data storage architecture that manages data as objects (as opposed to other storage architecture)

• file systems which manages data as files and fire hierarchy
• block storage which manages data as blocks within sectors and tracks

Storage Classes

(less cheaper)

• Standard(default)
• Intelligent Tiering
• Standard Infrequently Accessed (IA)
• One Zone IA
• Glacier
  • for cold data
    • the points : price(The lowest price), safety(99.9% durability) -> so it’s the best for cold data
• Glacier Deep Archive

(the cheapest)

1

Amazon S3 Glacier is a secure, durable, and extremely low-cost Amazon S3 storage class for data archiving and long-term backup. With S3 Glacier, customers can store their data cost effectively for months, years, or even decades.

** reference : https://bluese05.tistory.com/35

AWS is guarantee of 99.99% of availability, 11's durability.

Storage Classes Comparison

** reference : https://www.youtube.com/watch?v=Ia-UEYYR44s

Security

• Log files are generated and saved in a different bucket. (not the same bucket) (Logging per request can be turned on a bucket.)

• ACL(Access Control Lists
  • the way to control access simply.
• Bucket Policies
  • it is generally used more.
  • you have to write a JSON document policy
  • a lot more rich, complex rules.
  • if you are ever setting up static S3, you definitely have to use a bucket policy
  • there’s no fool proof.

Encryption

• SSE (Server Side Encryption)
  • SSE-AES : AES-256 (uses 256 bytes, very long encryption)
  • SSE-KMS
    • Key Management Service
    • it uses envelope encryption
    • so the key is then encrypted with another key
    • It’s either managed by AWS or managed by you the key itself.
  • SSE-C
    • C : Customer provided keys
    • it’s no option in default encryption
    • it’s just encrypting the files locally and then uploading them to S3.

Data Consistency

• Consistency

  Consistency is going to be different when you are overwriting files or deleting objects.

New Objects(PUTS)

• when you put data or write new data to S3 as a new object, it’s going to be read after write consistency.
• What that means is as soon as you upload it, you can immediately read the data and it’s going to be consistent.

Overwrite Objects(PUTS) / Delete Objects(DELETES)

• when it comes to overwriting and deleting objects, it’s going to take time for s3 to replicated to all those other AZs. (it only takes like a second or two to update)

• so if you were to immediately read the data, s3 may return to you an old copy.

AZ

• Availability Zone
• the logical building block that makes up an AWS Region.
• the isolated locations within data center regions from which public cloud services originate and operate.

CRR (Cross Region Replication)

• it provides higher durability in the case of disaster
• when you turn it on, we are going to specify a destination bucket in another region.
• and it’s going to automatically replicate those objects from the region source.
• it replicate to a bucket in another AWS account which is setting on another region.
• for it, you need to have versioning turned on in both the source and destination buckets.

Versioning

• what versioning does is it allows you to version your objects.

• it help you to prevent data lose
• and just keeping track of versions.
• if you put a new file which has the exact same key, it’s going to create a new version of it. and it’s goint to it a new version ID
• the idea is if you access this object, it’s always goint to pull the one from the top.
• and if you were to delete that object, now it’s going to access the previous one.

• SO it’s a really good way of protecting your data.

• if you do turn on S3 versioning, you cannot disable it after the fact. which means once it’s turned on, you cannot remove versioning from existing file.
• all you can do is suspend versioning.

Lifecycle Management

• what it does is it automates the process of moving objects to different storage classes or deleting them all together.
• it can be used together with versioning
• it can be applied to both current and previous versions

Transfer Acceleration

• it provides you fast and secure transfer of files over long distances between your end users and S3 bucket.

Presigned Urls

• it generate a url which authenticate you temporary access to an object to either upload or download object data.
• it is commonly used to provide access to private objects.
• You can use AWS CLI / AWS SDK to generate it. ** CLI : Commend Line Interface ** SDK : Software Development Kit

MFA Delete (Multi Factor Authentication)

• it makes you require an MFA code in order to delete object.
• there’s some limitations (caveats) around how you can use it or you can’t use MFA delete.
  • You must use AWS CLI in order to turn on MFA
  • You have to make sure version is turned on your bucket
  • Only the bucket owner logged in as the root user can delete objects from the bucket.
• This is going to be a really good way to ensure that files do not get deleted by act.

Important words

• stand for : 象徴する。代表
• replicate : duplicate, 複製する。
• take a look : 一度見る
• have a feature called ~
• compliancy reason : 遵守理由
• that is the concept
• caveats : 注意事項, 注意事项
• durability : 耐久性
• optimizing operational excellence : 운영 우수성의 최적화
• that kind of stuff
• versus : vs
• elasticity : 탄력성 (elasticity of demand)
• scalability : 확장성
• a bunch of ~ : 묶음, 다발
• pillar : 축

• nitty-gritty

  what is essential and basic

• be applicable to N

• you will be overwhelmed

  to overcome completely in mind or feeling

• foundational information

• resilient

  able to withstand or recover quickly from difficult conditions. strong and not easily damaged by being hit, stretched, or squeezed.

• reliable and resilient storage

  able to be trusted.