Compare AWS, Microsoft Azure and Google Cloud for Backup

Cloud users typically choose a multi-vendor approach for their data backup and storage tasks. Using different solutions leads to more flexibility and often allows to cut overall expenditures. Azure, AWS and Google are among world's top providers of cloud storage services, so it is important to know about their differences in order to choose a right vendor for each particular need.

Google Cloud Storage

Google Cloud Storage is a business-ready alternative for Google Drive.
There are three storage classes:

• Standard
• Durable Reduced Availability
• Nearline

The Standard class guarantees high availability which means that stored data and services can be accessed any time. It also guarantees low latency allowing to retrieve data within tens of milliseconds. The DRA class provides lower availability and Google recommends using it for batch processes and similar scenarios. The Nearline class provides even lower availability and higher latency so it is recommended to use it for prolonged storage of data which are not urgently needed.

All three classes allow to book redundant storage facilities and multi-regional options. All of them support data versioning on a bucket level. Server-side data encryption is performed for each scenario. Stored files can also be compressed on demand. Each bucket allows to configure lifecycle settings for objects stored there. Besides, the Nearline storage class users will be billed for data retrieval or their early deletion. More details on pricing can be found at Google's Pricing page.

               

Microsoft Azure Storage

MS Azure's cloud storage is included into its  basic package of services and is accessible from the main dashboard for any registered user.

Azure has two storage classes:

• Standard with two access tiers:
  • Hot - for frequently accessed data, 99.9% SLA.
  • Cool - less frequently accessed data at a lower storage cost, 99% SLA.
• Premium

Premium storage is currently available for Azure Virtual Machines only. It allows to support a virtual disk operating with high workloads. A user can choose from three disk sizes: 128 GB, 512 GB and 1,024 GB. Standard class supports several scenarios of data storage. One of them is Azure Blob storage. Blobs are objects containing binary, document or media data and can be used for storing backups. An Azure user can create as many blobs as needed, though their entire volume cannot exceed 500 TB.

When creating a backup in a Standard class storage, user has to select a replication level: that is, the number of copies of data which should be created and where they should be stored, physically. Several copies stored in different facilities are less likely to be affected by any kind of disaster. The following levels are available:

• Locally Redundant Storage - three copies within a single facility; provided at a discount;
• Zone Redundant Storage - three copies, two or three facilities in one or two different locations; it is more costly
• Geo-Redundant Storage - six copies in two different locations; the price is higher than for ZRS
• Read Access Geo-Redundant Storage - the same as the above plus the user has full access to the data stored at the secondary location; the price is higher than for GRS

With the new Blob storage accounts (available via the new Azure Resource Manager), users will now be able to choose cool access tier to store long lived, less frequently accessed data at a lower storage cost, while retaining the same high durability level. This also provides latency and throughput comparable to current block blob storage when accessed, so you can efficiently store archives, backups, media content, etc. A user can change the access tier for a Blob Storage Account anytime.

Both classes allow versioning management. Data encryption can be performed on the server- or client side. Stored objects' lifecycle is managed by selecting a retention policy for it. Data retrieval from Azure storage is free, other pricing details can be found here.

 

Amazon S3 Storage

Amazon's S3 is also available from the basic AWS dashboard and has three storage classes:

• Standard
• Standard - Infrequent Access
• Reduced Redundancy Storage
• Amazon Glacier

The Standard Class is a great option for urgent data because of its low latency and easy access to stored objects. The Infrequent Access class provides basically the same services but takes a data retrieval fee while its cost of storage per 1GB is lower. The Glacier class is designed for prolonged data archiving as it has higher latency, lower availability and data retrieval fee. All three classes ensure data encryption at server side and while transferring plus allow life cycle management and versioning. Pricing details can be found at S3 Pricing page.

Let us summarize the differences between these vendors:

Feature	Microsoft Azure	Amazon S3	Google Cloud Storage
Availability	99.9% for Standard class (99% for Cool tier)	99.9% for Standard and IA classes	99.9% for Standard class
Versioning	Supported	Supported	Supported
Lifecycle Management	Supported	Supported	Supported
Encryption	Server-side + on-demand client-side	Server-side + Server -side with user password	Server-side
Compression	Supported	Not supported	On-demand
Data Retrieval	Free	$0.01 per GB for IA and Glacier classes	$0.01 per GB for Nearline class

 

How Can CloudBerry Help

So, each of the vendors listed above has certain pros and cons, and sometimes it is worth using different backup services for different needs. CloudBerry's solutions are seamlessly integrated with all three of them, allowing you to utilize the following enhancements:

• Select any of these vendors and create a backup in a storage class of your choice directly from a CloudBerry UI.
• Perform compression of the stored data on your own, on the client side, so that your data not only required less storage capacity, but also took less traffic expenditures while transferring
• Add client-side encryption and SSL for data transfer for better security - you will be an only keeper of the password
• Transfer your data faster with CloudBerry's multipart/multithread upload
• When using Amazon Glacier, manage the peak retrieval rate of your data to cut overall retrieval costs