Virtual machines still comprise the core IT infrastructure of most government agencies, which poses a challenge for CIOs and CTOs who have as a major part of their strategies a need to find ways to utilize the cloud.
On-premises applications are often huge monoliths, their capabilities so interconnected that it’s difficult to make changes without impacting the whole. To take full advantage of the cloud, one option is to separate those monoliths into smaller, less dependent parts known as microservices.
However, that level of modernization can be costly and time-consuming and delay the benefits of cloud migration.
An alternate and equally viable option is to migrate the virtual machines that run these applications to the cloud as they are, to ‘lift-and-shift’ the entire monolith in its current state as the first phase in a cloud adoption strategy, followed by a modernization phase.
Which option is best? It depends. For decision-makers, the key questions to ask are: What to migrate, when to migrate and how to migrate?
Signs an Agency Should Migrate
There are signs that an agency should begin migrating to the cloud. One is when datacenter infrastructure is nearing its end of life. That is, when hardware agreements are ending and refreshes are imminent. Likewise, when software licenses are due for renewal, it may be time to consider migrating.
Other signs that an agency should migrate sooner rather than later include:
- When disaster recovery is costly or difficult to implement in the current environment
- When applications need to scale rapidly as a result of seasonal demand
- When the need to deliver new services is critical but infeasible with legacy infrastructure
To make data-driven decisions about when to migrate, agencies should compile an accurate and up-to-date overview of their current environment. This overview should capture current resource utilization, TCO information, application dependencies and timelines for software license renewals and hardware refreshes.
When to Lift-and-Shift (And When Not)
Re-architecting on-premises applications to make them cloud native can take time. It also requires employees with the skill sets and experience necessary to convert monolithic applications into microservices. When timelines and expertise constitute a significant challenge, it makes more sense to lift-and-shift virtual machines to the cloud.
On the other hand, legacy applications with complex dependencies on on-premises systems may not be ideal candidates for an immediate move to the cloud. The same holds for applications with compliance and security requirements for sensitive data that are difficult to achieve in the cloud.
It’s also worth noting that cloud migration can result in cost savings, but only if agencies plan, architect and deploy workloads in ways which ensure that these savings are realized.
Planning the Migration
A misconception about cloud migration is that the cloud provider manages and takes complete responsibility for a customer’s workloads. In reality, there are various cloud-delivery models — including IaaS, PaaS and SaaS — that determine to what extent the customer and cloud provider are responsible for managing OS and software updates, managing cloud infrastructure, configuring security policies and other tasks.
Agency technology leaders need to understand the subtle differences between on-premises and cloud technology paradigms as they move from physical and virtual infrastructure to distributed systems and managed services. This will help inform their overarching cloud strategy and serve as a basis for setting cloud adoption goals, measuring success and reducing the chance of failure.
Countdown to Migration
Migrating workloads involves the movement of large amounts of data between the agency’s environment and the cloud provider. This requires a large enough pipe and adequate bandwidth between the two locations to ensure migrations complete in an acceptable timeframe and that day-to-day activities are not impacted.
As a result, the actual migration of workloads to the cloud requires careful planning and tools that help accelerate the process while reducing the amount of downtime required for applications and services. Agencies should also define and test a process beforehand that allows workloads to revert back to on-premises systems in the event that a cloud migration fails.
Closing Thoughts
Every agency’s journey is different, but all face the same pressure to modernize. Customers expect the government to provide services with the same seamless delivery they’re used to when accessing commercial applications on their phones or in a web browser.
For some, that will mean significant upgrades to their datacenter infrastructure and for others, it means modernizing their applications on-premises before migrating to the cloud. But for many others, the simplest way to get started is to lift-and-shift their virtual machines to the cloud.
Want to learn more? Request a free cloud assessment today.
