Before we dig into the questions you should ask yourself when moving to the cloud, let me give you the tl;dr (too long; didn’t read) for this blog. YES, you should move to the cloud. But I posit you should continue reading this post to understand some of the thoughts and concerns you need to consider when making the decision.
Moving to the cloud is such a generic idea. It is the technical equivalent of asking my spouse if we should eat dinner. Of course, the answer is yes, but the questions now lead to what kind of food, atmosphere, budget, etc. do I need to consider as I plan my dinner.
Consider the following:
What are you moving to the Cloud?
There are a lot of use cases that can show the value you will get out of moving various services to a cloud provider. Maybe you’ve been hosting your email on an exchange server on your network and are finally coming to realize the amount you spend on licensing and maintenance, not to mention the ever-growing mailbox sizes, backups and additional network overhead from the entire process. Moving these to a provider like G Suite or Office 365 removes a lot of the overhead and allows you to focus on your core business.
On the other hand, moving custom applications the company has built which contain private data or proprietary processes can be a bit daunting. In truth, it’s a step in the right direction. More on that in a bit.
I once worked with a security consultant who told me the only way to keep my server from ever being hacked was to bury it a mile down under the Mojave Desert and hope nobody ever finds it. Nothing is impenetrable, but we should be looking for as many ways as possible to keep the doors locked and the windows shut. A lot of executives have heard horror stories of moving their stuff off premise, but with the correct configuration, your data is no less safe in the cloud than it was in a datacenter. You will also have the bonus of additional resources monitoring security as most cloud providers are already knee deep in monitoring (and possibly more mature given the depth that they deal with).
You can’t tell me that you can maintain long term on premise systems for less than what most cloud providers are offering. If you can, you are probably doing it wrong anyway. To really compare the two, you would need to have your VM replicated and ready to stand up in another data center at a moment’s notice. You should also look at the maintenance cost. Your internal resources are probably spending more time managing on premise compared to the time spent with the cloud version.
Also, if you have internal applications, you can most likely move them to containers or hosted applications. That’s where you will begin to feel the greatest savings. Gone are the days of patching your servers and managing VMs. Now your teams can focus on putting the applications together. Going one step further, you may even find that the services themselves have a cloud based version (think exchange servers) where you can migrate your users to a cloud based service that removes almost all the maintenance overhead (it will never go away completely).
On the flip side, if you are a company that depends on your own internal hardware to crunch and process information, then you would need to see if there is a cloud based option for what you are doing. If you are a large search company (Google) or running a bitcoin mining operation, you will probably get more value out of keeping those services in house.
What is your plan for the next 3 years? What about the next 5? As you begin to look out, you must keep in mind what your internal resource constraints are. You might think you have enough resources to manage for the next 12 months, but an event could occur that drives business up and you aren’t able to scale fast enough. Cloud providers make it easier to scale up, often at the flip of a switch. It will cost you, but if your business is growing that fast, your bill for these services is probably low on the level of concern.
How do you handle maintenance? Do you have a series of environments where you deploy updates and patches to ensure that by the time you get to patching production environments, you won’t have any surprises? What about hardware? Do you have an end of life on hardware and a plan to upgrade periodically? What about virus protection? Depending on how your move your services and what you move, much of this can fall off your plate.
Is performance an issue? Is your database too slow? Do you need to throw more resources at it but recognize the cost and demand is extreme? What about network performance? Are resources outside of your office (customers or employees) hitting a bottleneck getting into your network at a prime moment? Moving yourself to a pay per use method is going to allow you to open that connection as needed. If everything seems to come to a head at 4PM on a Wednesday, then quiets down by 5PM, it doesn’t make sense to increase your internal pipeline to make that single hour not bring your business to a halt.
I suppose I should amend my initial statement that everyone should move to the cloud. Chances are that you may have services or applications that don’t make sense moving or your infrastructure is such that it would be more work than its worth. You could even be planning to close your doors in a few months and don’t care about the long term success of your business. If that’s the case, I apologize for the suggestion. Otherwise, what are you waiting for? You should always be evaluating your systems to find out how to optimize your business. Ask yourself this question: “Am I in the business of running a server farm or am I in some other industry?”