Cloud computing has become an integral part of the IT sector. The days of struggling with complicated networking and on-premise server rooms are long gone. Thanks to cloud computing, services are now secure, reliable, and cost-effective.
When we talk of top cloud computing providers, there are 2 names that are ruling the markets right now- AWS and Google Cloud. Here, we are going to compare both of them, determining the pros and cons of both. Before we start with the comparison, you should have an understanding of the latest trends in the field of cloud computing.
In the sixth annual State of the Cloud Survey of the RightScale, where over 1,000 professionals were interviewed, there were some interesting findings that came out:
The report was published in 2016, and since then significant changes have occurred in the field of cloud computing. Hosting sites at AWS and Google Cloud has become fairly easy. There are multiple WordPress hosting providers allowing you to use the cloud without worrying about the technical aspects of cloud computing. Several large enterprises are investing in their engineers and employees and helping them gain certifications offered by cloud computing platforms:
Over the past couple of years, security and performance have significantly improved. This is because cloud computing providers have come up with new ways of securely hosting data and delivering it faster. All the traffic between the data centers is now encrypted by default.
When it comes to public cloud adoption, AWS is still the leader. The main reason behind this being that AWS was the first cloud computing service to be launched and has significantly shaped the cloud industry. However, other cloud computing providers like Google Cloud and Azure have seen significant growth too.
Let’s take an in-depth look at these two market leaders in cloud computing to help you select the best one for your organization.
With all the different solutions and services provided by the Google Cloud Platform, you will be able to use the same hardware and software infrastructure used by Google for its own products like Gmail and YouTube. Their first service, Google App Engine, was launched in 2008 in public purview. Here are some of their products:
According to the Chief Executive Officer of Google, Sundar Pichai, Google Cloud Platform is one of the top three priorities for the company. The annual run rate for the platform is over $8 billion.
A subsidiary of Amazon.com, this cloud computing service was launched in 2006. Since then, it has offered multiple solutions and services. Here are a few of their products:
There are some big brands that are using the service provided by the AWS Cloud including Netflix, Nasa, Lamborghini, Time Inc., Airbnb, Expedia, etc.
There are many services that are similar between Google Cloud and AWS. With so many products offered by both, we can’t compare them product wise. Instead, we will be covering them according to their compute instances, storage, networking, and billing features.
Let’s compare how both the providers handle their instances i.e., their virtual machines. For virtual machines, Google cloud uses KVM while AWS EC2 uses Xen. Both the technologies offer predefined configurations with a specified amount of network, RAM, and virtual CPU. However, Amazon EC2 refers to them as instance types while Google Compute Engine refers to them as machine types.
With AWS EC2, you can equip up to 3,904 GB of RAM and 128 vCPUs. For Google Compute Engine instances, you can equip 3,844 GB of RAM and 160 vCPUs. Google Cloud also allows departing from the predefined configuration and customizing your RAM and CPU resources for fitting your workload. There are other types including AWS EC2 Spot Instances and Google Cloud Preemptible VMs.
This is a very important consideration as it will directly impact the performance of your applications like max IOP per instance/volume, expected throughput (IO), and the ability of bursting capacity for short times. When comparing AWS and Google, there are two types of primary storage that need to be considered: object storage and block storage.
Block storage is the virtual disk volume that is used in conjunction with cloud-based virtual machines. AWS EC2 provides this with its Elastic Block Store (EBS) while Google Compute engine uses persistent disks.
Object Storage, also known as distributed object storage, are hosted services used to store and access a large number of blobs or binary data. Google Compute Engine uses Google Cloud Storage to provide this service while AWS uses the S3 service for this.
Apart from the above-mentioned, both the providers also allow the usage of disks locally attached to the physical machine that is used to run the instance. When compared to persistent disks, this local storage provides very low latency, very high input/output operations per second, and superior performance. You can even achieve several GBs of read and write speeds with this storage, which is incredibly huge. AWS EC2 calls them instance store volumes while Google Cloud refers to them as local SSDs. Google Cloud allows attaching local SSDs to any type of instance. In the case of AWS, only the X1, R3, M3, I3, I2, HI1, G2, F1, and C3 can support instance store volumes. In 2017, Google Cloud announced a price cut on local SSDs for preemptable and on-demand instances.
Both the providers use different partners and networks for interconnecting their data centers and delivering content to end users via ISPs. For accomplishing this, different products are used.
When it comes to Google Compute Engine instances, the achievable network capacity is based on your VM’s CPUs quantity. For peak performance, every core is provided with a 2 Gbits/second cap. Every core increases the network capabilities to a maximum of 16 Gbps for every virtual machine.
Amazon EC2 instances, for the large instance sizes, have a maximum bandwidth of 25 Gbps. 10 Gbps/second is the maximum speed for standard instances.
When you are comparing the network capabilities of both the providers, network latency plays a major part. When you are working with the business with visitors from a particular geographic location, latency is important. For example, if you have a website in Frankfurt and more than 90% of your customers are from Germany, you will benefit from placing the site on a server in Germany rather than placing it in Asia or the United States. This can make a difference of about 2 seconds. It includes other factors as well like TTFB, DNS, etc. Both, the AWS and Google Cloud, have multiple locations across the globe for you to choose from.
On a latency test conducted using Cloud Harmony that offers impartial, reliable, and objective analysis of the performance, 50 servers located around the globe were utilized. The results showed that Google Cloud offered better latency. But the test was run from a specific location. Different location can give different results. For measuring ping times and latency, you can try spinning up small instances on both the providers and running your own tests.
Both providers have a different approach for billing. Both of them also have a very complicated way of doing it. You can try checking out their monthly calculators:
Calculating this monthly amount is not an easy task. There are tools like Cloudability and reOptimize that are built entirely for helping you better understand your bills. Google Cloud Platform uses its BigQuery tool for providing estimated exports. AWS has a dashboard providing insights to your bill. However, both of the cloud platforms are working their best to reduce costs and making billing easier.
In September 2017, AWS announced per second billing. This works great for clients who are working on spinning up new instances and carry out a large amount of work in a short duration. After this, Google Cloud also launched the per second billing. This just shows the intense competition between the two where they are simultaneously launching new products.
If you are seriously invested in one of the platforms, they will provide you with various ways to save costs. Reserved Instances is one such way by which AWS EC2 offers a significant discount and when used in a particular availability zone, provides a capacity reservation. There are 3 types of reserved instances:
Google Cloud uses Committed Use Discounts to all the customers of Compute Engine. So basically, in return for discounted prices, you have to buy the committed use contracts. After analysis, it was found that on using the 1 year standard RI of AWS vs the 1 year committed use discount of Google, the Google’s environment cost 28 percent less than AWS. The 3 year program for both the discount types led to 35 percent less cost in Google environment as opposed to the AWS.
Both, the AWS and Google Cloud, have multiple community forums and documentation that can help you understand their services for free.
However, you will have to pay for instant support or assistance. Both of them have support plans. We strongly recommend that you read the fees involved in both before availing of the assistance services. Both of them offer unlimited number of billing and account support cases without any long-term contracts.
For Google, there are 3 levels of support available - Silver, Gold, and Platinum. The cheapest plan is the Silver one starting at $150/month. The Gold plan starts at $400/month. You will also be charged a product usage fee of minimum 9% which will decrease as your spend increases.
AWS provides 4 levels of support - Basic, Developer, Business, and Enterprise. The cheapest paid plan is the developer starting at $29 per month of 3% of your monthly usage. The Business plan starts at $100 per month along with 10% of product usage fees which will also decrease as the spend increases.
When it comes to monthly uptime percentage, both have SLAs providing at least 99.95%. For staying up to date with the incidents, you need to subscribe to their status page. However, both the providers have delayed updating their status dashboards.
With AWS, you have the advantage of having different machines within multiple availability zones per region. On the Google Cloud, the same machine per region might have all your instances. However, with Google Cloud you have the ability of live migrating the virtual machines which allow addressing issues like patching, updating and repairing without worrying about the machine reboots.
In the Clutch’s Second Annual Cloud Computing Survey, it was found that about 70% of professionals felt secure storing their data in the cloud than on their previous, on-premises legacy systems.
With Google Cloud Security, you get the benefit of a security model that has been developing over a period of 15 years and is securing products like Gmail, Search, etc. There are about 500 full-time security professionals employed by Google. It provides security features like:
AWS platform also has a security model with the following features:
From the above it is clear that both cloud computing providers have their pros and cons. Google Cloud has seen rapid global expansion over the past few years. It is also the one to go for if you favour speed and affordable pricing. AWS has been a long-standing name in the history of cloud computing. AWS started it all and is still being copied by other major players in the market. AWS redundancy, support and availability per region have helped it stay at the top. Rest assured, the constant battle between both the cloud providers will result in increased performance, more services and products, and lower prices benefitting hosting partners and customers. You can try the AWS Certification course for learning about all the services offered by AWS.
Your email address will not be published. Required fields are marked *
A solution architect is an AWS solutions Architect... Read More
AWS has dominated the field of cloud computing in ... Read More