10 things to consider before your Azure migration

Tools can help, but they are no magic wands Your Azure migration is about more than just VMs Tools are often part of a lift-and-shift strategy. It may seem tempting to lean toward tools that fast-track your workloads to Azure, but there are also some things to consider when deciding on the tools that you use. If you are primarily lifting and shifting existing workloads to Azure, Virtual Machines will likely make up the majority of your Azure environment, but there are multiple considerations beyond VMs that require strategy and prioritisation. Some other pieces to the Azure puzzle that require a closer look are: While these tools can help you migrate workloads rapidly, doing so often comes at the expense of developing a design better-optimized for Azure. Don’t be afraid to use the migration to Azure as an opportunity to undertake some vital housekeeping. »» Connectivity How will end users access workloads in Azure, and what will the end-user experience be like? Do you need ExpressRoute to support your connectivity and security goals, or can you achieve your goals with VPN connections? Second, the third-party tools cost money, and Microsoft offers many native Azure tools and services that can provide the same functions without spending extra money on third-party products. If you do decide to lift-and-shift, consider the Microsoft tools first. »» Storage Later in this paper, we will talk about storage redundancy options, but you also have options for storage performance. The most common option, Managed Storage, has both Premium and Standard tiers to support a variety of IOPS requirements. 5 Patching Changes Just in the past year alone, we have witnessed two major eye-opening security breaches (WannaCry and Spectre) that remind us of the need for a strong focus on security and understanding how the responsibility lies. Microsoft automatically patches Azure infrastructure, PaaS and SaaS services for you. The client is then only responsible for IaaS VM patching, which can be completed through Microsoft Operations Management Suite. 3 6 Designing for Solution Resiliency Resiliency is the combination of high availability and disaster recovery. Whatever your application needs, Azure has you covered from storage and compute resiliency through backup and disaster recovery solutions. Keep the following concepts in mind as you design your solutions: »» Azure storage comes in locally redundant, zone redundant and geo redundant storage. With locally redundant storage, your data is stored across servers in multiple physical racks in the same datacenter. Zone redundancy adds a second data center in the same geographical region, while geo-redundancy replicates the data to a datacenter in another region. Prices go up as the redundancy goes up, so be sure to pick the right level of redundancy for your workload. »» Azure Availability Sets keep the virtual machines in separate update and failure domains. For workloads that have application level high availability, this prevents either planned or unplanned downtime from impacting all nodes at the same time. For example, placing two nodes of an Always On SQL cluster in an availability set results in the two nodes living in different physical racks and having different update schedules. »» When it comes to backing up workloads in or to Azure, you have a myriad of options. MARS, the Microsoft Azure Recovery Service, provides easy backups for simple workloads that you want to backup directly to Azure. MABS, the Microsoft Azure Backup Server, is a free solution that supports disk-to-disk to cloud for more complex workloads like SQL. DPM, Data Protection Manager, Microsoft’s backup solution from the System Center Suite, can either be run in Azure or configured to backup to Azure. It has the most flexibility in terms of supported workloads and backup targets. In many cases, the best way to achieve your resiliency goals is to consider using Azure’s PaaS solutions like Azure SQL. With Elastic Database tools and built-in backups, Azure SQL can support demanding resiliency requirements natively. 7 Using Azure tools to control costs One of the most attractive facets to Azure is the flexibility; flexibility within Azure is the most powerful lever for controlling costs because you can easily right-size and reduce use when VMs are not needed and, in turn, streamline and optimise your spend. We suggest under-sizing your virtual machines, not only because this can save you money and because many times businesses tend to have over-sized virtual machines, but also because it is simpler to upsize, if needed, than downsize. Beyond planning for VM size, your purchasing model has a huge impact on costs, as reserved VM instances and hybrid-use benefit can significantly optimise Azure compute costs. Finally, as a final layer of cost control, you can utilise other Microsoft technologies like Power BI analytics dashboards, and New Signature’s own Cloud Management Portal as a self-service tool to gain insight into your usage and spend. 4
Please complete the form to gain access to this content