Optimizing Storage Costs With S3 Storage Classes

Read on Beehiiv | June 27, 2024

Welcome to The Cloud Economist!

Last week I discussed how to optimize costs on serverless workloads with services like AWS Lambda and Amazon SQS.

The curated articles shared valuable tips on using Lambda’s Power Tuning tool to balance performance and costs, as well as how ARM architectures make your Lambda functions more efficient and save you costs.

If you missed that you can read it here.

In last week’s question, I asked, “What is one way to save costs with Amazon SQS that doesn’t require any tradeoffs on performance?“

This can be done with long polling: setting a higher wait time in the ReceiveMessageWaitTimeSeconds parameter in SQS, can help lower costs by reducing the number of empty responses and unnecessary polling requests.

This week, let’s discuss how you can save money with the most used AWS service: Amazon S3. We’ll look at various techniques to optimize S3 costs including the different storage classes we can use.

Here are the best articles I’ve found on cloud cost savings this week, summarized.

Article 1

S3 Intelligent Tiering Pricing and Potential Cost Savings by Vantage Team

Amazon S3 Intelligent-Tiering is an automated storage service that optimizes cost by moving objects between different storage tiers based on access patterns.

It can lead to significant cost savings (up to 95%) by transitioning infrequently accessed data to cheaper storage classes like Archive or Deep Archive tiers.

However, S3 Intelligent-Tiering charges monitoring fees per object, and savings are only realized over time as objects transition between tiers, meaning that initial costs may increase before long-term savings are achieved.

Here’s a concise breakdown of the potential savings of the different S3 storage classes:

For a deeper dive on which storage class to use based on your use case, I recommend you check out the article (a few minutes read only).

Article 2

How to Make Sure You’re Not Overpaying AWS S3 by Tobias Mildenberger.

To avoid overpaying for AWS S3 storage, it’s essential to understand its pay-as-you-go pricing model, which charges for storage, data transfer, and requests.

You can reduce costs by optimizing stored data (e.g., compression), using S3 Lifecycle policies to transition data to cheaper storage, and committing to AWS Savings Plans for long-term savings.

Additionally, monitoring usage and considering alternative, lower-cost storage providers like Contabo can help further minimize expenses.

One Tip on Cloud Cost Savings

S3 Select allows you to retrieve only specific data from within a larger object stored in S3, using SQL-like queries. Instead of downloading the entire object, you can extract just the relevant parts, reducing data transfer and processing costs, which can significantly save both time and money.

(*it is now only available to AWS customers that have S3 Select before July 25, 2024)

This Week’s Question

How can you optimize costs when uploading multiple files to Amazon S3?

Send in your answers by replying or wait for next week for the answer.

Until next week.

The Cloud Economist

Uriel Bitton