Amazon RDS is an IAAS service

Amazon Web Services versus Microsoft Windows Azure

After Microsoft had to fight for on-premise market shares with providers such as Novell, Oracle, IBM or HP in the last few decades, a new giant has established itself in the public cloud with Amazon Web Services (AWS), which is now increasingly putting its feelers in the that reaches out to corporate customers. This market is predominantly dominated by Microsoft and holds enormous potential for providers in terms of cloud computing. Forecasts by Crisp Research show significant growth of 40 percent per year over the next few years, which means that sales in Germany alone should amount to 28 billion euros in 2018.

Amazon Web Services: IaaS market leader

Already at AWS re: Invent 2013, Amazon Web Services (AWS) started with some new services to attract corporate customers. After AWS established itself as a leading infrastructure provider and enabler for startups and new business models in the cloud, the Seattle company has been trying for some time to get a foothold in the lucrative corporate environment.


Amazon Web Services are currently the undisputed market leader in terms of public IaaS. As a division of, the company has been operating several data centers and so-called edge locations in nine regions around the world since 2006 in order to deliver data more quickly. With constant investments in its infrastructure and portfolio, Amazon shows how important technologies are for the company. Amazon uses its cloud infrastructure for all of its own offers and services.

In addition to the basic infrastructure services Amazon EC2 (Amazon Elastic Compute Engine, computing power) and Amazon S3 (Amazon Simple Storage Service, storage space), many other infrastructure-related services are available with which customers can use the infrastructure effectively. It pays off. According to estimates by Crisp Research, an average AWS customer uses around 11 to 12 AWS services to run their web application on the Amazon cloud infrastructure. AWS is releasing new services at a remarkable pace, thereby consolidating its attractiveness as a driver of innovation in the cloud market. Now that AWS has been able to name a large number of the world's successful startups as its customers, the next step is now to attract established companies to the cloud infrastructure.

However, the scale-out infrastructure poses a challenge when using the AWS cloud. If web applications are to be really highly available, they must first be developed for the AWS cloud. The start of a single EC2 instance is never enough to speak of a cloud-capable application. This requires a network of several EC2 instances over which the application runs in a distributed manner. If an instance fails, a new one is started; during a peak load, new instances are started up and then shut down again later.

The Amazon infrastructure

The AWS infrastructure currently spans nine different regions worldwide. Each region consists of at least two independent and isolated data centers, the so-called Availability Zones (AZ). To ensure the availability and redundancy of an application, a multi-AZ concept should be used.

AWS Service Portfolio: Amazon EC2

Amazon EC2 (Amazon Elastic Compute Cloud) is the top service from AWS and is used extensively by customers. EC2 is one of the first compute services in the cloud and has continued to develop since its availability in 2008. In particular, the geographical range, the different VM types, the support of different operating systems as well as the different types of reference and the ecosystem are currently the measure of all things in the cloud. Based on customer feedback, Amazon is constantly expanding its service with improvements and new functions. Amazon responded immediately to requests from corporate customers for better I / O throughput for Oracle and SAP workloads and provided new SSD instance types for this purpose. Business solutions from Microsoft, Oracle, Red Hat, SAP, IBM and other software providers are certified for operation on Amazon EC2. Customers can choose between On-Demand, Reserved and Spot Instances to process their workloads with the appropriate computing power as required.

Amazon S3

Amazon S3 (Amazon Simple Storage Service) is Amazon's storage service. It is used both by many customers for worldwide access to data in web applications and by companies for backing up data. According to Amazon, over two trillion objects were already stored in Amazon S3 in April 2013. In addition, there are around 1.1 million accesses per second via the API. Amazon S3 serves as the storage backbone for many other AWS services, including Elastic Block Store (EBS), Amazon Relational Database Service (RDS), Amazon Elastic MapReduce (EMR), Amazon CloudFront, and Amazon Storage Gateway.

Amazon Glacier, Storage Gateway, CloudFront

Amazon Glacier is an archiving service that can be used for storing and backing up compressed data and is based on Amazon S3. The Amazon Storage Gateway is a virtual appliance that companies can use to automatically transfer backups of their data to the Amazon cloud. Finally, Amazon CloudFront is the AWS Content Delivery Network, which offers a worldwide distribution of data via edge locations and is tightly integrated with Amazon S3.

Amazon Virtual Private Cloud (VPC)

Amazon Virtual Private Cloud (VPC) is mainly used by corporate customers who connect their on-premise infrastructures to the Amazon cloud. With VPC, an isolated area can be set up within the AWS PublicCloud in order to process more sensitive data. For this purpose, virtual networks are set up, which are adapted and controlled by the customer himself. Public servers are deployed within a public subnet. Back-end systems as well as application and database servers, on the other hand, are located within a private subnet that cannot be accessed publicly. In combination with AWS Direct Connect, which can be used to set up a dedicated connection between a company data center and an AWS region, Amazon VPC is the most secure way to operate a hybrid infrastructure environment in the AWS context. Customers from the financial sector, government agencies and healthcare companies in particular use this form of connectivity to transfer workloads to the public cloud.

Amazon Relational Database Service (RDS), Redshift

The Amazon Relational Database Service (RDS) is Amazon's relational database provided as a service. At the beginning only MySQL was supported. In the meantime, Oracle and Microsoft SQL can also be used. Amazon RDS supports high availability, Multi-AZ concepts and replication for read access and can be combined with Amazon VPC. Amazon Redshift is a data warehouse service that has high I / O performance and is intended for the analysis of large amounts of data. The service plays a central role in Amazon's strategy to bring companies to its cloud infrastructure.

Conclusion: Amazon Web Services

Amazon is the leading IaaS provider and serves the requirements of various customer groups. While AWS was primarily used by startups at the beginning, services such as Amazon Redshift, Amazon Storage Gateway, AWS Direct Connect, Amazon WorkSpaces, AWS CloudTrail and Amazon Kinesis are aimed particularly at corporate customers.

Despite all denials from AWS Senior Vice President Andy Jassy, ​​it is to be expected that Amazon will also have a say in the private cloud market in order to compete against providers such as Microsoft and VMware. However, Amazon still has some catching up to do here. Using "AWS Direct Connect" and "Amazon VPC" you can connect your own infrastructure to the Amazon public cloud on the network level. At the application and management level, however, there is a lack of powerful software for the on-premise side. With the "AWS Management Portal for vCenter", a plug-in for the VMware vCenter, with which virtual machines can be transferred to the Amazon Cloud, AWS has now taken a first step. This is the right strategy for increasing its penetration among corporate customers. After all, VMware is still widespread in many IT infrastructures and AWS has the potential to attack here. However, the tool is currently still a rudimentary solution with which virtual machines can only be imported into the Amazon cloud. The administration and the automated transfer of workloads are not possible with it.

The requirements that Amazon must meet for companies vary depending on the country and use case. European customers are usually more cautious with data storage and prefer to store the data in their own country. Crisp Research has already advised more than one customer who was technically convinced of the Amazon cloud, but storing the data in Ireland was out of the question.

In some cases, however, the lack of ease of use also plays a role. This means that a company does not want to (re) develop its existing application or website for the Amazon infrastructure. The reasons for this are a lack of time and the knowledge to implement something like this. In addition, this would often extend the time to market. Both can be attributed to the fact that it is relatively complex to achieve scalability and availability on the Amazon Web Services. It's not just a couple of API calls. Rather, the entire architecture must be aligned with the AWS cloud. In Amazon's case, it is particularly the horizontal scaling (scale-out) that makes this necessary. On the other hand, companies would rather prefer vertical scaling (scale-up). In this way, you could migrate your existing system 1: 1 and would not have to start over in order to achieve success in the cloud.