Cloud computing has increase on its usage in big steps. We are a time where to have a cloud infrastructure reliable, secure and cost-effective is pretty much easy. At the end of the day we want to see our money well invested.
At 4Geeks we have been used Google, AWS and Azure, but I want to focus on Google and Amazon ( I like them more ). My mission is to show benefits of them and to know how and when to implement them in a project.
This is totally personal, and we can discuss here, but I lost less when using Google Cloud Console, there are limited options here, and trust me, this is good for it. Console of Amazon AWS has lots of stuff which we really don’t need and they only get us lost a little bit. Most of the times we don’t need that quantity of features, and honestly they only get higher the learning curve.
We enjoy to work with cleanest sites where the UI is pretty straight forward and Google does that.
Billing and Pricing
I will like to go first with the free trial, those days where our wallet breaths and we desire to keep it forever, but it won’t, so let’s go ahead and check wich one offers the best free trial, so our team can test and confirm.
Google offers $300 as credit that can be used in the first 12 months when you activate your account, I find this really cool because you can figure out how much your environment will cost. There is also a free forever set up where you can run an small application and the cost will be $0, cool right ? Well, the instance is not the big thing, for free you can run an instance with 0.2 CPU and 0.6 GB of RAM with 30GB of disk and 5GB of cloud storage.
Amazon on the other corner, doesn’t have credit to be spend, but They have a free trial for 12 months within a specific set up. So basically you can run 750 hours/month of a small 1 CPU/1GB RAM instance with 30GB disk storage, 750 hours/month of a database instance and 5GB of cloud storage (similar to Google), but the trial doesn’t limit to those products, you can check the entire list on the link above, but you can find RDS, Load Balancing, CloudFront under the expense list of free products.
Now, when the the fairy tale ends, We have to pay, but, how much ? which one is expensive ?
So Google is like 32% less expensive than AWS, using the example above. There are other features we may use to our projects like Google App Engine, Google Kubernetes or Amazon Elastic Beanstalk, but for our porpuse with that setups is enough to see that Google is cheaper utilizing the exactly same setup.
I really like Google’s free trial because you can manage your “money” and you can realize quickly how much your setup will be, also, when go live, you will save money utilizing Google for sure.
Google invest on talent about security measures. They create events internal and external to keep sharing security knowledge. Google has in internal audit team that searches for security laws changes around the world, for example, the new compilance GDPR. Personally I give a lot of value to the issue of having specialized teams, that gives value to the company and assures me that the people behind the products are indicated, they show the team to the world.
All communication and data transfer are encrypted by default, when we send data to Google they encrypt the data, and we the data is sent to Data Centers too. Google servers uses string security protocols as TLS to secure communication between Google and Customer device, when we use Google Load Balancer we trigger those communications protocols, also, there is an option to use Google Cloud VPN, which establishes IPSec virtual private networks.Google encrypt all persistent disk are FDE (Full Disk Encryption) that protects data at rest.
Something really cool is that Google brings us some the capability to use third party certifications, bringing value to our systems. We found it really interest un 4Geeks, since we use PCI compliance for our payment gateway.
Google has IAM implementation which creates specific access for users to the entire Google Cloud Platform. This is cool because we can handle satisfactory the access to our infrastructure avoiding any unwanted access.
Amazon VPC is a great tool to handle firewalls, so you can create virtual network and have total control of the access to your infrastructure, this is very useful to increase security to our systems. All transit communications are encrypted over TLS protocol.
Amazon provides also a flexible key managment so you can basically handle keys easily to the different encryption process, but AWS can handle that as well for you. So you have all keys centralized, we don’t tend to lost them anymore.
There is a service similar to Google IAM called AWS IAM if you are thinking which one was the first with the name or the idea, well, it was Amazon. So IAM work the same both platforms, but Amazon go further and provides Multi-Factor authentication, so, besides having restricted access you can ensure the veracity of the logged user.
Finally they do some compliance services such as PCI, ISO, HIPAA and SOC.
Why Do I go with Kubernetes? Well … this is an excellent of container orchestration. We suggest more and more clients every day to use Kubernetes as the container handler. So, our team had to start doing comparison between which platform provides the best environments and pricing for Kubernetes development.
Google is always ahead of Kubernetes updates since the do it natively on all their software, so, they know it well. With GKE(Google Container Engine) you can easily start doing Kubernetes with a smaller learning curve than Amazon EC2 container Service probably It’ll require more ground work.
This is a topic that has been growing fast, more and more production environments are migrating to serverless. At 4geeks we have had implemente a couple of projects with serverless approaches.
We moved from create server infrastructure from scratch, which it can take lot of weeks, lot of money, lot of headaches to hace services that creates infrastructure for us like Kubernetes and Puppet. So, it’s simple to deploy an app, but what about if you want to run functions when requested, not taking care of your environment, updates and all the ground work to our a project top and running, well … serverless.
I took it from here. I found that post really interesting. Both of them provides a really stable solution in serverless function. I think that AWS is good because you have more options in languages. The finality of this sever less comparison is to take a better understanding of what serverless, also knows that there is options taking care of it. We will talk about serverless a little bit deeper in the future, so I invite you to subscribe if you want to know about it.
All projects works in different ways they need different resources and need different platforms and approaches, that’s why we believe there is no ultimate tool or platform to be used, whenever you need some help of our dedicated DevOps and architecture team you can reach out to us.
We really love all DevOps topics, so I, personally, will present a webinar: Importance of DevOps on a successful project. I will go ahead and discuss why we need to focus on the culture of DevOps and the importance to choose the right toolset for our project. You can always visit the webinar under our site.