Different users of Amazon S3 can receive different bills for similar data. It can happen because each user works with information in their own way and that results in additional costs.
In this article, we will explain Amazon S3 pricing structure and storage costs estimation.
Amazon S3 Regions
Regions are one of the fundamental factors in Amazon S3, and the cost of one gigabyte depends on data location. For example, you will pay $0.03/GB for standard storage in North Virginia, Oregon and Singapore and more for data located in Sao Paulo.
When estimating Amazon S3 costs, make sure you know average region prices. Below we shared rates for Standard storage class in each AWS region.
Amazon Web Services offers six discount tiers applied depending on the total amount of data stored.
Simply speaking, the more data you store in Amazon S3, the less money you pay for each gigabyte. Discount tiers don’t apply to Standard I/A storage class and vary through different AWS regions. We’ve prepared a comparison table to illustrate the impact of volume discounts on Standard storage class in each AWS region.
|N. Virginia, Oregon, Ireland, Ohio||N.California||Tokyo, Sydney, Seoul||Frankfurt||Sao Paulo|
|1 TB – 50 TB||$0.023 per GB||$0.026 per GB||$0.025 per GB||$0.0245 per GB||$0.0405 per GB|
|50 TB – 500 TB||$0.022 per GB||$0.025 per GB||$0.024 per GB||$0.0235 per GB||$0.039 per GB|
|> 500 TB||$0.021 per GB||$0.024 per GB||$0.023 per GB||$0.0225 per GB||$0.0037 per GB|
Note: The presented regions and volume rates are valid on 4/26/2018
If you have to store 52 TB (53248 GB) of data in Standard storage class in N.Virginia region, you will pay the following amount per month:
$0.022 * 53248 = $1 171.46
Remember to consider volume discounts each time you estimate the cost of Amazon S3.
Requests and Data Transfer Fees
Each time you access the data, you send requests:
- LIST requests that return the current list of stored objects.
- PUT requests to upload new objects and other actions.
The number of requests depends on the number of objects you affect. For example, while uploading 12,000 files, you will make 12,000 PUT queries. The price for each request also varies between different AWS regions.
|N. Virginia||N.California||Tokyo||Seoul||Frankfurt||Sao Paulo|
|PUT, COPY, POST, or LIST Requests (per 1000 requests)|
(per 1000 requests)
|GET and all other Requests (per 10,000 requests)|
(per 10,000 requests)
Note: The rates are valid on 4/26/2018
Getting back to the example with 52 TB stored, the cost of your requests will be added to the monthly bill. Let’s imagine we need to upload the same amount of data in 1,240,000 files (or objects) to Standard storage in N. Virginia region. Then we decide to retrieve half of them. It means that we will make 1,240,000 PUT and 620,000 GET requests.
Here are cost calculations:
$0.0220 * 53248 = $1 171.46 (Storage)
$0.005 * (1240000/1000) = $6.2 (PUT requests)
$0.0004 * (620000/10000) = $0.02 (GET requests)
Total: $1 177.68
Standard I/A storage class is distinguished by a specific type of requests: Lifecycle Transition Requests. Bills for this type are separate and usually it is used to configure a corresponding backup retention policy and object lifecycle.
The data transfer fee is charged for each gigabyte transferred inside or outside Amazon S3. Like the volume discount system, this fee has discount tiers from the first gigabyte up to 5 PB of data transferred via the Internet. Additional charges are applied to transfers between AWS regions.
When you sign up to AWS for the first time, you get started with AWS Free Usage Tier. The offer lasts for one year and includes 5 GB of Amazon S3 storage with the following monthly limitations:
- 20,000 GET Requests
- 2,000 PUT Requests
- 15GB of Data Transfer Out
If you stay in this tier, AWS won’t charge anything unless the amount of requests or the storage capacity is exceeded. Free Usage Tier is the best way to test if Amazon S3 fits your requirements.
Storage classes and Amazon Glacier
AWS Amazon S3 has four storage classes: Standard, Standard I/A, Standard One-Zone I/A and Reduced Redundancy Storage (RRS). Each one was designed for different purposes.
- Standard is an example of a “hot” storage class for the highest durability and availability.
- Standard I/A is “cool” storage for infrequently accessed data with reduced availability.
- RRS is the same as Standard but has less redundancy. So you may lose one of 10,000 files per year.
There’s also an archival storage called Amazon Glacier, which may be at some point considered as a low-cost storage class for “cold” data. Data is stored in archives and the retrieval process can take more than four hours. Amazon Glacier has a separate pricing model. Read more in Amazon Glacier Retrieval Pricing Explained article.
Each class has its own unique pricing and only two of them (Standard and RRS) offer volume discounts. It’s very important to understand which storage class is better in each case to estimate future costs. Read more about AWS S3 Storage classes pricing and features in our blog post.
In this article we’ve covered the most important concepts in Amazon S3 pricing that influence the monthly bill:
- AWS Regions Pricing
- Volume discounts
- Requests and Data Transfer Pricing
- Storage Classes specifics