- diagram
- Almost all workloads generate some data, and often the data should be stored in files (or databases)
- files generated by applications
- ex. web app) user uploads, invoices, transformed images, etc
- ex. analytics app) generated pdf reports
- file archiving
- ex. business) account related, legal docs, etc
Different kinds of file storage
- AWS has one service for every kind of file storage
- diagram
Block storage
- EBS (Elastic Block Store)
- Attaching a (virtual) hard drive to a server
- Unformatted hard drive you attach
- Once you attach it, you have to format & structure it before you can use it
- For EC2 instances, you attach a block storage device, and from in the instance you mount the hard drive and format it.
- Then you can create a custom structure and use it like any other hard drive to create a file system
Object storage
- S3 (Simple Storage Service)
- The idea
- Store & retrieve files, focus on the individual files themselves
- Any kind of file w/ any size
- Doesn’t care about underlying system/hard drive/file system where the file gets stored
- no custom structure/file system
File system
- EFS (Elastic File System)
- FXs
- A file system lets you organize, store, and share files.
- Think of it like a shared folder in the cloud that multiple computers (EC2 instances) can read and write to simultaneously.
- Network file system where you can attach multiple devices (ex. ec2 instances)
- You don’t want to care about the underlying hard drive
- Preformatted file system, n that system you can create custom structures & files
- X care about hard drive or objects, care about the file system
- You can attach to different services, then you can store files as well
- related: 🐧Linux
EBS vs EFS
EBS vs EFS vs S3
- diagram
