AWS offers a significant array of storage options that adjust to your needs. So, one thing that is critical at this time, is to understand your exact use case. Let me break it down for you for simplicity. Say compute optimized

storage, for example. This is going to include several services that are designed to be attached to virtual machines in order to capture data as it's being generated. The most specific example of this would be EBS or Elastic Block Storage. There are other options available such as EFS or Elastic File System, and FSX. All of these three options are tied to several performance tiers to cause and also, durability. Again, it all boils down to your use case. Object-based storage. This type of storage is going to be specific to storing objects, that is, data artifacts that are not fully dependent on compute. For example, a backup of your production database may not need to be used at all for months or years. So, there's no need to associate this large backup file with a virtual machine, it's just an object. The undisputed champion in this arena is S3. Generally speaking, S3's performance for real-time needs is poor because it's not designed for streaming or real-time management of information. So, this would not be your first choice if you're looking for speed.

That  said, when it comes to durability, S3 offers 11 nines of durability. Translation, your data is replicated both physically, across physical hard drives, and geographically, across several data centers. So, you can rest assure your data is safe. Another strong factor around S3 is its cost. It offers several tiers of costs that you can take advantage of. Just understand that cost tiers will affect your retrieval times and durability. Hybrid computing storage. If cloud storage wasn't complex enough, you can always add one more layer of complexity. In this case, I'm talking about the ability to use and store data locally while at the same time, safe-keeping it in the cloud. Fortunately, AWS offers devices that can be attached to your on-premises network as just another network appliance, and also an option to do the same thing as an on-premises virtual machine. The name of this service is Storage Gateway, and it makes transferring data between on-premises and AWS completely seamless and transparent. Data transport storage. At some point, you may be faced with a data migration from Point A to Point B that involves more than a handful of terabytes of information. In this case, transmitting the data over wires could be impractical.

It could take days or weeks while at the same time slowing down your corporate network for this purpose. This is when you have to consider data transport storage option from AWS. Devices such as AWS Snowball or Snowmobile, are designed for those edge cases from a few dozen terabytes all the way to petabyte scale. And believe me, with 4K video right here with us and 8K videos slowly coming around, getting into the hundreds of terabytes will soon be a normal occurrence for many of us. Here's a list of use cases that you can refer to and hopefully find one that is similar to yours. Continuous backups that need to be readily available from on-prem to the cloud. The recommended option would be storage gateway appliance or virtual appliance. Cross-region backups with low Recovery Time Objective; S3 would be the recommended option here. Data that needs to be accessed frequently but is not dependent on performance; again, S3 is the recommended options, but you also have the option of using EFS, which you can mount directly to your instances. Storage for real-time data intensive computations.

You can use EC2 ephemeral storage or provisioned I/O EBS volumes. Storage for application binaries, operating systems, or just general purpose; standard EBS volumes. One-time data transfer; you can use Snowball or Snowmobile. Data sharing across virtual machines and Docker containers; you can use instance-mounted EFS volumes.