ra2 studio - Fotolia

Cloud migration planning: Assess workloads now to avoid trouble later

Organizations that take the time to assess workloads, identify dependencies and educate themselves on cloud security tools can set a smooth path for a public cloud deployment.

Editor's note: Cloud migration, whether a lift-and-shift project or a more sophisticated application refactoring initiative, is a task more customers are pursuing as they seek the advantages of public cloud computing. But without careful cloud migration planning, those migration ventures may not result in the hoped-for benefits.

IT service providers, however, can play a role in keeping their customers' on track with their public cloud deployment efforts. In this article, John O'Shaughnessy, a senior consultant with the cloud and data center transformation division at Insight, a provider of computer hardware, software, cloud solutions and IT services, discusses the measures organizations should take to make sure their workloads are cloud-ready.

You often gain just as much wisdom from someone's past failures as you do from their successes. This is also true for organizations looking to migrate one or more workloads to the cloud. Here, we look at some lessons learned from a variety of companies' cloud migration planning mishaps.

Cloud migration planning: Workload assessment

In the last few years' rush to cloud, many enterprise and midrange companies have adopted a cloud-first mentality. For some, this means accelerated migration of workloads to a public cloud.

With any workload, performing initial groundwork is critical to determine the best underlying platform for migration.

Yet, not all workloads are created equal, nor are they equally best-suited for public cloud. With any workload, performing initial groundwork is critical to determine the best underlying platform for migration. This includes assessing both the technical and business needs of the workload, then deciding which underlying platform offers the best fit.

The right platform might turn out to be a public cloud, a private cloud or, even, a hybrid cloud environment. For aging legacy workloads, significant rework, rearchitecture, consolidation or retirement might also be on the table, as well as more traditional "lift-and-shift" options.

Not performing this type of due diligence prior to migration can have costly repercussions. This includes one of the more drastic: the need to migrate a workload back out of a public cloud environment. In 2017, IDG Research conducted a survey commissioned by Datalink, the cloud and data center division of Insight Enterprises. Over half of the respondents (52%) reported having to move one or more workloads back from a public cloud to an on-premises model.

This type of real-world experience has made organizations more measured today in their approach to cloud migration. Seventy five percent of IT leaders in the survey said they were now more cautious than they had been a year ago when deciding whether or not to migrate specific workloads to a public cloud environment.

Getting the details right: Identifying dependencies

Assessing technical and business needs of a workload is a good start to ensure you make the right platform choice. But, here, as with any migration, getting the details right really matters.

Some legacy applications have many dependencies and data flows. All of these interconnecting parts can be tough to identify and detangle before the workload can be made "cloud-ready." Even then, some may never get there without investing first in significant rearchitecture. 

Other workloads may have strict compliance and security needs that also may not make public cloud a good fit.

Then there are seemingly benign workloads with web front ends that look to scale easily. These are often considered ideal for cloud migration. While many may be, careful vetting is still needed before moving even these types of workloads to a public cloud environment.

One case in point is identified in the white paper, "Public Cloud Workload Migration: 9 Common Mistakes to Avoid." Here, the organization's upper management wanted to migrate several application workloads to the cloud. The workloads had preexisting web front ends and seemed easily scalable. Why not do so?

During the initial assessment period, the back-end database for these workloads was found to also be critical to hundreds of other applications in the organization. All of those apps needed ongoing access and communication with the database. Further, it was discovered that as many as 50 apps comprised a single workload, with each app similarly dependent on the same database.

Could this workload go to the cloud as it stood? Not likely. Application owners were not surprised by the assessment's outcome. This particular legacy architecture was too monolithic and tangled to make such workloads a good, short-term candidate for cloud migration. Only after the assessment, however, did company management realize that such workloads (in their current state) could not be easily migrated. 

Understanding security, implementing governance

Many mistakes identified in the white paper can be avoided with the right upfront, internal research regarding each workload. They can also be avoided by using best practices to assess the right workload or platform alignment.

Additionally, thorough understanding of the tools and techniques for managing security and implementing governance in the cloud will be key to a successful public cloud deployment.

A lack of defined objectives, outcomes and measurements for post-migration success also can trip up organizations' cloud migration efforts.

Ultimately, investing time conducting due diligence on the front end ensures that organizations lay the groundwork for a smooth transition to public cloud and gain the benefits public cloud has to offer.

For more coverage on cloud migration planning, check out this Essential Guide on migration strategy. And learn how public cloud deployment trends have affected channel partner investments.

Dig Deeper on MSP technology services

Cloud Computing
Data Management
Business Analytics