Cloud Basics: GCP, Azure and AWS

The landscape of information has changed dramatically over the last decade, not only in terms of sheer volume but also in terms of the way that we store it. Once upon a time businesses and individuals maintained their own servers, keeping information locally. Remember floppy discs and thumb drives? How about hard drives that crashed and lost everything? The cloud eliminates all of that and offers users the immediate access from anywhere, with a low risk of loss.

Big Players

There are big players in cloud storage. Amazon, Microsoft and Google are dominant, massive entities in information technology. There’s a good reason for that — cloud storage means trusting someone with your precious information. These three are trusted companies and users feel confident will be there for the long haul. Cloud technology offers a tremendous amount of versatility and innovation potential, but at the same time it requires a level of trust in the platform that’s holding hard-earned and vital information. No one wants to think that their info is sitting in some guy’s basement, so startups aren’t the name of the game here.

Early in the process (think 2006), Amazon rolled out Amazon Web Services (AWS) as a data storage platform. Microsoft was next on the block with Azure in 2008, and Google Cloud Platform (GCP) debuted in 2010.

These three are ready to take a piece of the growing dollar signs that are associated with cloud storage and computing. While AWS has a head start on the other two, that doesn’t mean that Google or Microsoft are out of the race.

Let’s break down some major differences between GCP, AWS and Azure.

Networking, Storage and Speed

These are the basic functions of the cloud, the reason that we’re here. Everyone has something special to offer here, so it’s important to understand them before you make a choice.

DNS web service is a unique networking service provided by AWS (at least among these three). In addition it offers the same services as Azure. Amazon uses virtual networks that it calls Virtual Private Clouds (VPC), which is equivalent to Microsoft’s VNET.

Though Amazon has just one terabyte of block storage and is in the middle of the pack in terms of speed, its cloud offerings definitely make up for it in other ways. With a whopping instance overture of 7 families and 38 types plus elastic load balancing, AWS continues to widen its breadth of storage options.

AWS Computing

• Regions
• Zones
• Instance Types — 38
• Instance Families — 7

AWS Storage & Databases

• Ephemeral (Temporary)
• EBS Block Storage
• S3 Instance storage
• RDS Relational DB
• Glacier Archiving
• Dynamo DB, EMR, Kenesis, Redshift

AWS Networking

• VPC
• Public IP
• Hybrid Cloud
• Route 53
• Firewall/ACL

Azure has the most basic networking offerings of the three — including public IP access, hybrid cloud services and firewall protection. Microsoft calls its virtual network VNET, though unlike Google’s subnets it can’t be broken down into smaller pieces.

As for block storage, Azure presents just 1 terabyte for users, with instances amounting to 4 families and 33 types. In addition, Microsoft’s platform clocks in as the slowest among the three compared here.

Azure Computing

• Regions
• Instance Types — 33
• Instance Families — 4

Azure Storage & Databases

• Temporary Storage — D Drive
• Page Blobs Block Storage
• Block Blobs and Files Instance storage
• Relational DBs
• No Archiving
• Windows Azure Table HD Insight

Azure Networking

• VNet
• Public IP
• Hybrid Cloud
• Firewall/ACL

Offering a subnet VPN, Google has individual Google Compute Engines that can each be set up independently. Public IP addresses can be accepted on Google, and firewalls can be applied to each subnet. If you’re looking for an incredibly versatile networking platform, GCP is definitely your best bet among the three. The downside is that GCP doesn’t offer hybrid cloud networking, which can be a big problem.

Google has the big guns when it comes to storage — offering up to 10 terabytes. In terms of speed, GCP is also the fastest, in part because it’s the only major cloud service to rely on its own regional fiber network. In terms of instances, Google offers 4 families and 18 types, similar to Azure.

GCP Computing

• Regions
• Zones
• Instance Types — 18
• Instance Families — 4

GCP Storage & Databases

• Ephemeral (Temporary)
• Persistent Disk Block Storage
• Google Cloud Storage Instance storage
• Google Cloud SQL Relational DB
• Nearline Archiving
• Cloud Datastore, Big Query, Hadloop

GCP Networking

• Subnet
• Public IP
• Firewall/ACL

Services

Here’s where we get to some real divergence — services.

● AWS

With ten years in the business, Amazon has had plenty of time to create a broad and deep library of services that can fit the needs of almost any business. The company created targeted tools for everything from streaming to managed directories to databases to remote desktops and more. What’s more is that Amazon continues to develop improved services for its users that are fully integrated into its platform.

● Azure

Azure recently increased their service offerings. Microsoft offers a mature set of analytics products as well as solid business tools for tracking. As always, Microsoft is a leader in hardware integration services, with Hyper-V container technology rolling out with Windows Server 2016.

● GCP

Google is lacking considerably when compared to Amazon as far as services offered for cloud users. So far the company focuses mainly on IaaS and PaaS related services, though just as with all of these platforms there is anticipation of expansion.

Cloud Pricing Models

Price is crucial when it comes to cloud services, in part because the cloud becomes an integrated part of any business. Choosing the right plan means getting one that has pricing that works not just for now, but for the long term. There are several models of cloud pricing and it’s important to understand each of them before signing on the dotted line.

● AWS

Amazon offers three different ways to pay for cloud services. Spot payment offers you the chance to bid for the lowest possible rate, but only on extra capacity. For reserved payment, you pay upfront for use over the course of 1 to 3 years. With AWS on demand pricing you pay for what you use on the back end with no upfront costs. All cloud service costs are per hour and rounded up.

AWS Pricing

• Per hour & rounded up
• On demand, spot, reserved

● Azure

Azure has on demand pricing for cloud services, rounding up in minute increments. They also utilize short commitments, think one to three months, which allow customers to pay upfront and at a discounted rate.

Azure Pricing

• Per hour & rounded up that are monthly or prepaid
• On demand, short term commitments that are monthly or prepaid

● GCP

Google has recently rolled out sustained use pricing in addition to on demand pricing for its cloud services. This innovation is simpler than the structure created by AWS and will allow for discounts for on demand users. Like Azure, GCP rounds up data usage per minute, with a minimum of ten minutes.

GCP Pricing

• Per hour & rounded up with a 10 minute minimum
• On demand & sustained use

Overall Picture for the Cloud

It all comes down to details. Microsoft offers the least amount of flexibility, but also offers hardware integration that’s not present in the other top cloud services. Google has tremendous networking flexibility and software integration through its business apps, but is still developing its cloud services. Amazon has the longest record in the business, with the deepest library of resources and the most variety in pricing.

The bottom line is that all three of these services are highly ranked across metrics. No matter which one you go with, you’re going to get great support and a massive infrastructure to back you up and secure your data. There really is no wrong decision here, unless you count one that isn’t well informed.

Has your business integrated with a cloud service yet? Not sure how to do it? Contact us. We’d love to help you out!