5 Best Practices for Becoming an AWS Migrations Competency Partner
Mission achieved the AWS Migration Competency. Learn more about their recommended best practices to attain this important AWS competency.
Let’s talk about why customers are choosing AWS specifically for Microsoft workloads. There are several hyperscale cloud providers available in the market today including Microsoft themselves (Azure). But when it comes to why customers are primarily choosing AWS for these workloads, it really comes down to a few key factors as discussed below.
So, recognizing that AWS is unbeatable when it comes to Windows workloads, let’s consider some different migration options:
Next we get to re-hosting, re-platforming and re-factoring based on your business strategy. AWS also recommends you consider your business drivers, so that the strategy for re-hosting, re-platforming and re-factoring align very closely with your ‘lift and shift’ optimization steps.
If you are going to re-host and get into the cloud in a lightweight way, you get deep familiarity, full control, and reliability. You are not changing a lot but you are getting some of the initial administration experience with AWS. You can get full control over the environment and you can have full access to all the capabilities of SQL Server.
But once you are there, if your business driver changes to getting more optimized architecture, you can re-platform. That involves using things like the Elastic Beanstalk and Directory Service, and you can get a lot more automated patching and backups resulting in improved high availability. And then full re-factoring, where you finally start realizing the full benefits of public cloud, like integration with cloud ecosystem and major benefits to TCO. Licensing becomes a lot more transparent and on-demand. And you have a lot of access to migration tools and services with AWS.
SQL Server on AWS Exhibited 2X+ Better Price/ Performance than Azure
Let’s discuss a specific use case here about SQL Server. We work a lot with the customers that have large SQL server workloads. They ask us whether SQL Servers work better on Azure or AWS. An independent research firm, ZK Research, did an extensive study in 2018 on performance and TCO for SQL Server on both AWS and Azure, and they discovered two findings..
First, they compared based on cost per billion transactions per month. This is a kind of measure of the cost per performance. On AWS, the price-performance is a little over 2x compared to Azure.
Second, they compared based on performance benefits. Many people think that instance types across different clouds that look the same are actually the same, but they really are not. This is a common misconception. When you benchmark and compare the actual performance of different instance sizes that price-performance gap is even greater. So, if you look at AWS’ equivalent instance size compared to an Azure instance, you are actually getting near 3x performance improvement.
With this research in mind, let’s discuss two of Mission’s recent case studies that will help demonstrate the real-world benefits of migrating your Microsoft Windows workloads to AWS.
The first case study is VBO Tickets. VBO Tickets provides an online software platform that helps event producers to create, manage, and sell tickets online. In order to support a rapidly growing customer base, VBO identified the need to design their next-gen architecture and engaged with Mission to do so.
The main issue was that their existing environment was not designed for scalability, and traffic spikes were causing intermittent downtime for their customers. VBO’s main application was running on a single albeit large Windows server instance that also had the database running alongside it. Furthermore, the versions of Windows and SQL Server were nearing their End of Life (EOL) also about to become Out of Support. So VBO Tickets was in need of an architecture that could scale their variable traffic requirements as well as ability for them to develop and deploy new functionality in the quickest manner possible.
For the solution, Mission designed a completely new architecture using auto-scaling and load balancing that would allow VBO to scale the environment up and down accordingly with demand. This allows the company to insulate itself from the variable traffic spikes that were common to its business. We were able to decouple the database from the application and migrate it to RDS while upgrading both Windows and SQL Server to the latest versions at that time. Additionally, an entirely new Continuous Integration and Continuous Delivery (CI/CD) pipeline was created using Jenkins Powershell scripts and Git to manage automated releases and pave the way for a new Blue/Green deployment model where downtime could be minimized and the rollback process simplified.
The second case study is ATCC. ATCC maintains the largest and most diverse collection of biological materials in the world. After many years of being hosted at the same provider, ATCC was given a three-month notice that their data center provider would be shutting down. This started the clock on a race to migrate their entire infrastructure out of the data center into their new home which was to be AWS. This is never a fun circumstance for anyone. ATCC couldn’t afford any downtime because a huge portion of their business was through their e-commerce website
The solution to their accelerated timeline was that we elected to leverage CloudEndure, which hadn’t been purchased by AWS yet, to migrate those servers while also providing disaster recovery services post-migration. So, using that tool, Mission was able to migrate ATCC’s entire environment in a single day, albeit after several trial runs. This migration allowed us the time, post-migration, to troubleshoot any lingering issues and actually fine-tune and dial-in performance. In the end, ATCC was able to continue with their e-commerce operations with zero unplanned interruptions in service.
The time to take action is now. End of support is coming and in some cases, it has come, for several Windows workloads. SQL Server 2008 reached End of Life (EOL) in July 2019 and Windows Server 2008 will be reaching EOL in January 2020.
Mission can work with you to upgrade and modernize your Windows Workloads, and help you work with any licensing ramifications in doing so. Reach out to us if you have any questions about migrating to the newest MS software versions or if you need any assistance with your cloud journey.