Amazon S3 Review

Over the course of this Amazon S3 review, we’ll mention a few of those third-party clients. If you’d like a jump-start, give our best file-transfer clients guide a read. For more advanced needs, our CloudBerry Backup review and Storage Made Easy review highlight two of our favorite S3-compatible options.

Coming up, we’ll break down the confusing costs of Amazon S3 and look at its impressive server network. We’ll also touch on user experience, security and support to help you make sure S3 is the best cloud IaaS provider for your needs.

Alternatives for Amazon S3

Strengths & Weaknesses

Pricing

Understanding Amazon S3’s pricing structure can take some work. Not including Amazon Glacier, the company has three different storage tiers and bills differently by region. Like many IaaS solutions, the company also charges for both storage and usage. For some, there are API call charges to worry about, too.

The three Amazon S3 storage tiers are called Standard, Standard-Infrequent Access (Standard-IA) and One-Zone-Infrequent Access (One-Zone-IA).

Amazon S3 Standard provides hot storage, a term for storage that will be frequently accessed. Choose this if you intend to use Amazon S3 for cloud storage with a third-party client such as Storage Made Easy.

The two IA plans are designed for less routine access, making them better for online backup. Both Standard and Standard-IA store data in multiple regions for resiliency, while One-Zone-IA keeps data in a single region.

Amazon Glacier, which you can read about in our Amazon Glacier review, is designed for long-term storage of files that you rarely need to access.

Storage rates for S3 Standard are much higher than either IA plan but usage rates are much lower. To make things simple, we’ll look separately at costs for storage and usage. You can also use Amazon’s storage calculator to estimate costs.

Storage Rates

Amazon bills per gigabyte and per month. There are no minimum storage requirements and you can store as large a volume of data or as many objects as you require. There is a limit on the size of individual objects stored, which is 5TB.

For Standard storage, there are three different rates depending on how much you store, while the IA plans have one rate. Those rates are listed for the Ohio region in the table, below.

Storage rates vary by region, but not by much. For example, data stored in northern California has a Standard rate of 2.6 cents and the rate in Frankfurt is 2.45 cents.

The storage costs are reasonable for a network of Amazon’s size, and the company has lowered prices considerably in recent years to compete with other IaaS options. While more than Microsoft Azure hot storage by about half a cent per GB, S3 undercuts Rackspace Cloud Files by over 75 percent (read our Rackspace Cloud Files review).

If you want to try Amazon S3 first, sign up for a free 5GB plan, which is good for one year.

For much cheaper rates, you could go with Wasabi or Backblaze B2, but neither has the server infrastructure you’ll get with Amazon.

Usage Rates

Data uploads and downloads are free for S3 Standard. Both IA plans have a flat 1 cent per gigabyte rate for retrieval.

Free retrieval for hot storage data provides a cost benefit over Azure, which charges 1.84 cents per gigabyte for the first 50TB retrieved. The IA plan rates for retrieval are identical to those Microsoft charges for Azure “cool” storage.

For those who intend to use S3 for application development, there are charges for API calls like get, select, put and copy. You can find those detailed on the S3 pricing page.

Server Network

Amazon S3’s network infrastructure is a big reason the service maintains its position as the IaaS market leader. Though not as large as the Azure network, file copy speeds to and from the S3 network and multi-region storage are pluses.

Amazon has four S3 data centers in the U.S., located in northern Virginia, Ohio, Northern California and Oregon. There’s another North American facility in Canada and one in South America, The other centers are in the E.U. (four) and Asia-Pacific (six).

That’s 16 data centers, which is one more than Google Cloud, but Azure outpaces both by a mile with 52 server facilities.

Ease of Use

Before you can upload files to Amazon S3, you have to create storage buckets. Buckets are file containers used to simplify management, since otherwise you’d be storing every file in one place, whether for backup, cloud storage or some other purpose.

To add new buckets, log in to the Amazon AWS console and click “S3.”

The linked page will display buckets you’ve already created. There’s also a blue “create bucket” button you can use to add new ones.

Each bucket must be given a name that’s unique across all Amazon S3 accounts, not just yours. You’ll also need to pick a server region for your new bucket and set various features like file versioning, log creation, tags and encryption. To speed things up, you can choose to copy the settings from an already-created bucket, if you have one.

With your bucket created, you can start storing files. While the S3 interface lets you upload single files to a bucket, there’s no option to move files in bulk. For that, you’ll want to use a simple file-uploader like Cyberduck or a more feature-packed client like CloudBerry Backup.

In order to pair a third-party client with Amazon S3, you’ll need an API key. Retrieve this by clicking your name near the top of the interface and selecting “security credentials” from the drop-down menu. Then click the “access keys” header.

You’ll need two credentials to establish a connection: the key ID and the key itself. The key portion can only be generated once, so make sure you write it down or download the key file when given the chance.

We detail how to use that key to establish a connection with CloudBerry Backup in our how to backup with Amazon S3 guide. The steps are similar for establishing connections with other third-party clients, whether Duplicati, Storage Made Easy, Mountain Duck or another.

While using Amazon S3 can prove a little tricky at first since its buried in the larger AWS ecosystem, once you know what links to click, setting up and managing buckets should be easy.

Security

Amazon takes many steps to ensure S3 data security, including an option to set default encryption for storage buckets. That’s a nice addition since many IaaS providers only provide in-transit protection and put server-side encryption on the user’s shoulders.

The encryption protocol is the standard AES, used by many cloud services, set to 256-bits. You can also opt to use AWS-KMS, which is Amazon’s key management system. Amazon Key Management Service makes it easier to create and destroy encryptions keys, as well as log access.

Because weak passwords pose one of the more commonly exploited security holes, Amazon also supports multi-factor authentication. With MFA turned on, you’ll be required to enter an additional security code when logging in. This code gets delivered to your mobile phone by text. You can also use an MFA device like a key fob or smartphone app.

To ensure server facility integrity, Amazon undergoes third-party audits, limits access based on the principle of least privilege, controls access points using detection systems and multi-factor authentication and maintains 24/7 surveillance, among other measures.

Facilities are also secured against environmental threats like fires, floods and earthquakes using automatic sensors, responsive equipment and data redundancy measures in case the worst happens.

We won’t go into all of the safeguards Amazon has in place to protect its data centers, but they’re considerable, which doesn’t surprise us, considering that server facilities perform the backbone of Amazon’s core business.

While we have plenty of security concerns regarding Amazon’s cloud storage solution, Amazon Drive, S3 looks strong as far as security measures taken. There have been numerous reports of S3 breaches, but those tend to be the result of careless missteps by the users.

On the other hand, as the market leader, Amazon S3 no doubt offers a more tempting target for hackers and the like, so there is that to take into consideration.

Support

Good support for Amazon S3 requires an AWS support plan. There is a free, basic support plan that provides 24/7 access to customer service, but doesn’t include technical support, meaning it will only do for the odd question and, of course, sales and billing support.

The cheapest paid support option costs $29 a month and gets you tech support during business hours and only by email. Round-the-clock technical support, including both live chat and phone access, starts at $100 a month.

For simple questions, Amazon’s online knowledgebase and support forum should suffice, both of which are free to use. The knowledgebase can be searched and has both articles and tutorial videos. The support forum, as you might expect of a popular IaaS, is active and in most cases, you won’t be left waiting long for help.

Final Thoughts

It’s hard to fault users for flocking to Amazon S3. For an IaaS provider with its network and features, the storage and usage rates charged are more than reasonable, even if it can take several hours of head scratching to figure out how much it’s all going to cost.

Whether or not it’s truly the best Iaas available, despite leading the market, is a matter of debate. While second-place Microsoft Azure still trails Amazon S3 by 20 percentage points, it’s gaining, and it provides lower costs and a larger network. Cheaper options like Backblaze B2 and Wasabi are also on the rise.

While there’s a lot to like with Amazon S3, its days reigning atop the IaaS market might be numbered.

Feel free to chime in with your thoughts on S3’s place in the IaaS market in the comments below, and thanks for reading.