The cloud era is gaining greater speed, and market projections look very healthy for years to come. This growth and the enthusiasm it’s creating for cloud technology is pushing organizations who, until very recently, were resistant to start their own cloud initiatives. This post is about how your organization can migrate the smart way with Azure cloud.
Unfortunately, many firms think of a cloud-migration effort as just another IT project (similar to what was previously done to install new hardware and software). As a result, they are making costly missteps for three key reasons:
1. Lack of training
2. Not using cloud architecture best practices
3. Not taking cost control and monitoring seriously from the start
Consider the following scenario:
Susan is the IT director for Spry Maneuvers, a successful, mid-level logistics firm. SpryM specializes in moving difficult-to-ship freight (such as power plant gas turbines) across the globe.
SpryM is heavily dependent upon a complex database system that sprawls across a large number of virtual machines (VMs) hosted on-premise:
The business is growing, so there is always a need for more storage capacity, more computing power, and greater performance from VSphere’s production and development environments.
Susan is an experienced and knowledgeable manager, and she knows how to address this need through budgeting, hardware provisioning, software licensing, patching, and all the other items required to engineer and maintain a busy data center.
Recently, however, the CIO gave Susan a new directive: For the coming fiscal year, there will be no budget for further on-premises data center investment. Every new initiative must be built on Azure and every existing asset must be moved there too, ASAP.
Under pressure, Susan and her team decided to lift and shift the database farm to Azure by building VMs and simply moving the workloads onto them, just as they’ve done countless times before within their data center:
This seemed easy enough, and after a period of adjustment to the Azure portal interface, SpryM’s IT team began building more and more VMs and moving workloads to over-provisioned F-series virtual machines.
Everything seemed fine until the bill (for tens of thousands for only a few weeks of run time) came due and end-users began complaining about performance issues.
The CIO wondered:
- Wasn’t running things in Azure supposed to be cheaper?
- Wasn’t the cloud supposed to be faster and more efficient?
- How could my team let me down?
SpryM made several critical (and all too common) mistakes from the start:
1. They assumed that the cloud was nothing more than an extension of their data center, as currently configured.
2. Staff requests for training (from those who recognized the challenge) fell on deaf ears—too expensive not necessary.
3. The cloud project was too ambitious. There was no modest proof of concept in which a low impact workload was relocated to Azure to help IT gain experience and confidence
4. No one monitored costs.
How could they have done things better?
Cloud Academy offers excellent courses on Azure such as Architecting Microsoft Azure Solutions–70-534 Certification Preparation.
This learning path covers the following key areas:
- Designing Azure Resource Manager (ARM) networking
- Securing resources
- Designing an application storage and data access strategy
- Designing advanced applications
- Designing Azure Web and Mobile Apps
- Designing a management, monitoring, and business continuity strategy
- Architecting an Azure Compute infrastructure
These are the skills that SpryM’s IT team needed to make the right choices. In addition to courseware for individuals, Cloud Academy offers team-optimized training that could have put SpryM on the correct path.
Azure Solutions Architecture
Microsoft provides a wealth of guidance to help you intelligently and effectively apply the platform’s offerings to your needs. This should be one of the first places you visit and examine when you’re building with Azure:
Azure Pricing Calculator
SpryM’s IT team was unprepared for the high runtime cost of the type of VM that they chose. This embarrassing surprise was avoidable—they just needed to use the Azure Pricing Calculator:
Using the Pricing Calculator, they would’ve seen that, when compared to running SQL Server on an always-on virtual machine, using an Azure SQL database is very cost effective.
As you move workloads to Azure—whether you eagerly volunteered or were pushed by management—keep the combination of training, cloud-native solutions, and pricing foremost in mind and you’ll greatly increase your odds for success.