A Practical Guide to Cloud Migration and Moving Production Workloads

Assessing Your Production Workload

As part of a practical guide to cloud migration, and before migrating your production workload to the cloud, it’s crucial to assess your current infrastructure and workloads, as well as your perceived future demand if you know it, and build a migration plan. Initially, it’s best to start by identifying the specific applications and services that are essential for your production environment, as this gives a baseline of where the business currently sits. Consider factors such as data storage, processing power, and network bandwidth requirements.

Additionally, evaluate the security and compliance needs of your production workload. Cloud providers offer various security features, but you must ensure they align with your organisation’s regulatory requirements and data protection policies.

Choosing the Right Cloud Service Provider and Tools

Selecting the right cloud service provider is a critical decision in your migration journey. Three major players dominate the cloud industry: Amazon Web Services (AWS), Microsoft Azure, and Google Cloud Platform (GCP) – in this case we will be looking at Microsoft Azure. Each provider has its strengths and offers a wide range of services to meet different business needs depending on the requirement so be sure to evaluate all options.

All the providers have their own migration tools from Amazon’s Migration Evaluator and AWS Migration Acceleration Program (MAP) through to Google’s Cloud Migration Center. In this particular blog, we will be focusing on Microsoft Azure.

Azure Migration Tools

When considering Microsoft Azure as the cloud provider, utilising a tool called Azure Migrate can help to simplify the migration, helping to both modernise and optimise your service for Azure. The tool helps to plan the initial stages, such as the discovery, assessment, and sizing requirements – covering servers including SQL and Web Apps by looking at on-premises servers and workloads running on VMware and Hyper-V.

Azure Migrate can also drill down in assessing SQL server instances and databases assisting in the migration to SQL Server on Azure VM, Azure SQL Managed Instance or Azure SQL Database. This is on top of Web Applications to an Azure App Service and Azure Kubernetes, and through to Virtual Desktops and migrating large amounts of Data using Azure Data Box. Note that the tool provides a guide for the workloads and the business may wish to migrate said workload to a different type of service – for example, the tool may say to migrate a virtual machine with a SQL database to an Azure Virtual machine with SQL server, but instead you might decide to proceed with an Azure SQL Managed Instance.

In a practical guide to cloud migration, we would say it is also worth noting that the sizing guide estimates the size of the VMs or SQL instances in Azure prior to migration alongside estimated costs for running in Azure to provide a cost comparison. It also helps to provide a dependency analysis to identify cross-server that rely on one another and optimisation strategies for moving interdependent workloads.

Planning Your Migration

With a practical guide to cloud migration, a well-thought-out migration plan is essential for a successful transition to the cloud, so before ploughing ahead with the migration, there are some key steps to get the ball rolling. We have discussed the initial Azure Migration tool already but following a few key areas will help to lay out the foundations for a well thought out plan to be created. Here are some critical areas to consider:

• Inventory and Assessment: Document your existing infrastructure, applications, and data. Assess dependencies and performance requirements to determine which workloads are suitable for migration. The Azure Migrate tool can help here!
• Select Migration Method: Decide whether to use a rehosting (lift and shift), refactoring (rearchitecting), or re-platforming approach based on your workload’s complexity and goals. Note that a lift and shift approach may be the first step to get the organisation to the cloud, prior to rearchitecting or re-platforming the environment – as these areas could extend the time to achieve a cloud migration.
• Data Migration: Plan how you’ll transfer your data to the cloud, some businesses have the luxury of having a Microsoft Azure ExpressRoute in place but don’t be put off if you don’t, it will just need some further planning as to how to achieve it. Some ways could be to involve data transfer services, such as Azure Data box, or alternatively using backup and restore methods, or data replication, say over a weekend or other period of time. Critical data that is constantly changing may require some further thought.
• Testing: Create a testing environment in the cloud to validate your workload’s functionality once provisioned and with some or all data migrated to create an initial testing environment. This allows all teams to test and validate its efficacy and provide feedback as to whether the platform works the way it should. Analysing the performance in this environment allows the business to confirm before migrating production data.
  • • One of the key areas here is to provide a guide of what will be tested when the production workload is migrated to the cloud, whether this be a check list or a tick box exercise. Some businesses may even do their testing as if they were a customer to ensure that the experience is seamless without hiccups.
• Security and Compliance: Whether you have implemented security best practices in your current environment, it is essential to ensure that these are migrated or in place within the cloud environment. If not, then following best practice guides will help to introduce better compliance and security during the migration process. It’s always best to have them in place prior to moving to production rather than implementing them post migration which may introduce issues.
• Monitoring and Optimisation: Set up monitoring and management tools to track your workload’s performance in the cloud, particularly in the testing stage, to understand demand and performance. Once the environment is in production then having the ability to monitor but continuously optimise resource usage to control costs is key to ensuring an overall successful cloud migration.
• Backup and Disaster Recovery: When migrating to a cloud environment, most businesses will have a planned backup and disaster recovery run sheet on their current on-premise environment so replicating or potentially improving on this when in the cloud is key. Establish agreed business backup and disaster recovery mechanisms to safeguard your data and applications in the cloud. Tools such as Azure backup or Azure Site Recovery will help to meet these goals, but also enable the ability to backup locally in country or in new territories to protect against Azure availability zone issues whether on a Geo or Zonal level.

Executing the Migration

As you reach the next stage in the plan to migration, taking into consideration the pre-requisites and plan above if documented, then the migration process can start to take place. Agreeing the stages of the migration is vital to business operations and department approval, for example some potential non-critical workloads could be migrated first to ensure least impact to customers, whilst critical workloads that could impact customers may take a bit longer to execute – say migrating a file share to a virtual machine or Azure File Shares (impact felt within the business) vs migrating a critical database workload (impact felt by customers/suppliers).

This typically involves the following steps:

• Provision Resources: Create the necessary cloud resources set forth in the Azure Migrate tool, including virtual machines, storage, and networking components. Note that the Azure Migrate tool is there as a guide, so if you believe a different resource or service is better suited then proceed with the agreed business need.
• Data Migration: Transfer your data to the cloud using the chosen data migration method, some of these have been listed above. Be sure to coordinate as part of the plan whether to stage the migration of data over time factoring in dependencies, such as migrating user file data vs SQL database data.
• Application Deployment: Deploy your applications and services to the cloud environment, with the foundation of the environment built it is then up to each relevant department to configure and deploy their applications into the cloud prior to the next phase of deployment.
• Testing: Conduct thorough testing to ensure that your production workload operates as expected in the cloud environment. With the applications now in situ, it is up to each department to run through their testing process which hopefully will have been created as part of the planning phase (this could take some time depending on the complexity). Once all departments have performed their testing, there may be some teething or alterations that need to take place to ensure a seamless process of staff and/or customers/suppliers.
• Monitoring and Optimisation: Continuously monitor performance, security, and costs. As detailed in the planning phase it is key to have delegated resources who cater for the ongoing optimisation of the platform, not only for cost efficiency, but also for workload performance. Over time demand may increase, or even decrease, and adjustments may need to be made to optimise the cloud infrastructure.
• Training and Documentation: Provide training for your team on managing and operating in the cloud incorporating monitoring and optimisation tasks. Following the hopefully successful deployment and migration, it is critical to ensure comprehensive documentation for reference to avoid, not only infrastructure sprawl, but also knowledge transfer between departments. New IT or other staff members may need to understand the workings of the environment quickly, so a well-documented architecture streamlines this transfer.

Post-Migration Considerations

Once your production workload is successfully running in the cloud, many businesses tend to not be vigilant on demand within their industry or changes to cloud providers offerings – so it’s crucial to remain aware of the ability to adapt the environment to introduce new tools and/or services to increase performance or cost efficiency. Many cloud providers continue to introduce new services that could significantly improve performance or allow the ability to reengineer the infrastructure.

It continues to be best practice to regularly review and update your cloud infrastructure to accommodate evolving business and/or sector requirements and technology advancements – that said, IT is ever changing and so should your business. Additionally, consider implementing cloud-native services if the business believes it would help improve their position and introduce practices to further leverage some of the other benefits of the cloud.