As an MSP we hear the same story at least once a week from a current or prospective customer, “Our board has mandated our use of public cloud.” If you haven’t already heard it, don’t be surprised when you do. There are multiple benefits of moving to the public cloud, but developing a clear strategy and successfully making the migration are critical to realizing those benefits. So how do we do that? Having significant experience in this area, these are the top 4 keys to a successful migration we see today:
I recently spoke with a customer who wanted to walk through a public cloud migration. He really understood how critical this was to his organization’s success. Among many other items, we discussed moving his tier 2 applications first to ensure the security and compliance aspects could be met on these applications before moving on to his critical ones. This is always my first suggestion to anyone considering moving to the public cloud. Plan out your migrations by application or site and move the least critical first. You can then work your way up to your most critical, which helps iron out any possible challenges with the migration and/or with performance after the migration.
One of the common misconceptions for those who have not yet worked with the public cloud is that, since the operating systems are “like-for-like” in both environments, everything should be a standard move. This is incorrect. Yes, file structures and settings, such as IIS or the amount of VCPU and RAM can be identical in both; however, the virtualization mechanisms, networking, load balancing tools, etc. will most likely be different. If it is a small environment with one machine, there won’t be too many challenges, but even then, you have to consider how networking will be handled. The key is identifying those differences and solving for them before migrating your application or site over is absolutely essential.To do this successfully you need experts who have been there and done it before. Whether you want to hire someone in house or bring on an MSP, make sure you bring in an expert to help you. It’s worth every dollar spent in avoiding pitfalls.
I cannot stress enough, the importance of testing the new environment in every way possible. Creating test data and scenarios, performing load testing with realistic volumes, and testing any applicable High Availability (HA) scenarios is the most critical step. Too many companies have made the mistake of running a basic test to ensure it works, but after migrating the site or application they realize they did not come close to putting it through the proper paces through testing. This then creates performance issues and in many cases, even full-on downtime. My boss likes to operate with the policy, “measure 3 times, cut once” to ensure we get things right before executing. This is the same in testing. Test at least 3 times before migrating. Trust me, your stakeholders will thank you for it when the migration goes smoothly.
The final step in any successful migration is understanding the tools you will use for migrating the data and the timing to get it right. Today so many applications or sites are business critical 24×7, yet we all know there is need for occasional maintenance with limited down time. A migration of environments is absolutely one of those times; in fact, it is nearly impossible to avoid. When migrating, it’s important to understand how much down time you can handle when you convert the environments, and then choose the proper tool to meet your need. If you have large amounts of data it may take weeks for the initial sync. You may need to do nightly or hourly syncs until you can complete the cutover. Can the tool you selected handle that, and does it support your database or file structure type of the data you are migrating? If so, you can then determine how long the initial sync will take, how often you need to sync after it is complete, and how long the last sync will take to complete once you take the system down. This will allow you to plan for the appropriate maintenance window, along with the appropriate time to ensure you have a successful migration. While at the same time, minimizing the downtime impact to your business and customers.
A successful migration to the public cloud can be complex. Each of the 4 items outlined above have importance in the overall process. If any one of them is skipped, it could lead to a migration failure that could be very difficult to recover from. When you break the message down to its core, it aligns with most lifecycle planning models: Plan, Prepare, Test, Execute. Each one has a piece of planning, preparing, testing and executing within them, but overall this aligns with a simple four step process for success. If you follow this process along with some of the key factors outlined in each step then you should be ready to lead your company into the public cloud.
Contegix can also help you with your migration and make sure it happens smoothly. Contact us today.