Migrating your datacenter to an IaaS platform like Azure is a big step for small businesses. There are many benefits to moving to the cloud – like increased productivity, better agility and decreased costs – but getting there is a daunting process.
In this guide, we’ll break down the specific steps involved in each phase to guarantee a successful migration.
Breaking Down Azure Migration Phases
Microsoft recommends a four-step migration process for migrating to Azure:
Discover: Catalog your software and workloads
Assess: Categorize applications and workloads
Target: Identify the destination(s) for each of your workloads
Migrate: Make the actual move
Let’s take a closer look the four Azure migration phases.
Discovery involves identifying all existing workloads and applications in your infrastructure so you can prepare them for migration. It’s an extensive and tedious process, but critical to success.
Missed applications and workloads can become headaches later on, so you’ll want to make sure your application inventory is complete and up-to-date.
Review these areas during the preparation phase:
To maintain the same datacenter performance, security and stability while managing costs, analyze your on-premises workloads in your existing virtual or physical environment and compare them to equivalent resources in Azure. Be sure to address networking requirements, how many subnets you’ll need to provision, and whether you’ll provide your own DNS servers via Active Directory.
Purchasing new storage every time you reach capacity is a constant burden. There are a few types of Azure storage to consider depending on the nature of your data.
- Standard vs. premium: Regular Azure storage has a certain IOPS maximum for each virtual disk. Premium storage delivers high-performance, low-latency disk support for virtual machines with input/output-intensive workloads.
- Hot vs. cold: How you store your data in Azure depends on how often users access it. A multi-temperature data management solution will help you conserve costs. Hot data requires fast storage, while data that is rarely accessed (cold data) is stored on the slowest storage.
The cloud gives you instant access to computing resources. When planning, you should look into Azure Autoscale. The autoscaling feature dynamically scales applications to meet changing performance requirements.
Azure virtual machines will also give you more control over your computing environment. An Azure VM gives you the flexibility of virtualization without having to buy and maintain costly physical hardware.
Once you have a better understanding of Azure products and how they fit into your migration strategy, it’s time to evaluate your existing infrastructure. Here are some tools to help:
This tool automatically inspects your on-premises environment, whether physical or virtualized, and provides a checklist for moving your workloads to the cloud. After the assessment, the tool generates a report detailing the workload attributes/configuration that are ready to move and what requires further investigation before moving. The report also provides additional resources to resolve issues and prepare the workload for a move to Azure.
The MAP Toolkit is an agentless inventory, assessment and reporting tool that securely assesses IT environments for various platform migrations including Windows 8.1, Windows 7, Office 2013, Office 2010, Office 365, Windows Server 2012 and Windows 2012 R2, SQL Server 2014, Hyper-V, Microsoft Private Cloud Fast Track and Azure. Since version 8.0, MAP Toolkit can assess environments and provide readiness information for both physical and virtualized workloads migrating to Azure.
Both of these tools can help you carefully review and document all applications, workloads and processes you currently use, including:
Current Infrastructure: Map your virtual and physical system configurations to an equivalent Azure instance. Evaluate specs like CPUs, disk size and storage demand.
Current Network Architecture and Capacity: Assessing your network architecture and capacity will help you evaluate bandwidth to replicate changes made on virtual machines. Use a capacity planning tool or bandwidth assessment tool to determine whether replicating a virtual machine would kill your network.
Performance Requirements: You need to know what IOPS you’ll require to avoid lags and maintain the same performance in your new Azure environment.
High Availability/Resilience Requirements: You need a system that will function in the event of failure. Thoroughly document your disaster recovery processes, resiliency configurations and recovery time objectives to ensure your data can be restored easily in your new environment.
Maintenance Process: Once you move to Azure, what maintenance steps need to happen to continue running effectively? Determine how your maintenance process will need to change in the new cloud environment.
Now that you’ve audited your existing environment, it’s time to map out how to get your servers in Azure.
Have questions about mapping your existing workloads in Azure?
The three likeliest targets for your workloads are:
- Microsoft Azure
- A Cloud OS Network
- Office 365
You’ll likely migrate productivity and communication-related workloads to Office 365. This may include moving email to Microsoft Exchange Online, document management to SharePoint Online and moving instant messaging, voice, video and shared application communications to Skype for Business Online (now Microsoft Teams).
Factors like speed, ease of migration, cost and desired functionality will inform the cloud destination for your workloads. For instance, websites would benefit most from the speed of Azure data centers, elasticity of the storage, processing power and memory. These factors help keep sites responsive even during peak demand.
Virtual machines are another component to consider. VMs are on the other side of an Internet connection and subject to the unpredictable nature of Internet latency. Most companies prefer to migrate non-critical VMs to less expensive cloud resources. You’ll want to put VMs in the cloud that won’t be adversely impacted by latency and don’t require frequent connections to other resources.
Operating systems are another consideration. Is the workload you wish to migrate running an Azure-supported operating system?
Make sure the VM you’re migrating doesn’t exceed 32 cores, or the maximum 448GB allowable memory.
Still have workloads running on physical servers? The terms for converting physical server workloads into a VM aren’t always straightforward. Working with a proven partner can streamline the process.
Now that you’ve audited and prepared your existing workloads and applications, you’re ready to migrate to Azure.
You can spend significant time reading up on best practices, studying available tools and preparing for the trial-and-error inherent in any new pursuit. But the potential savings are not worth the risk involved in attempting a complex cloud migration without the right expertise.