The cloud in all its variations has truly revolutionized business technology. It’s empowered smaller businesses to have the tools and features that only enterprises had previously enjoyed, and it’s allowed us to provide flexibility and access to employees everywhere. However, it’s not necessarily a one-size-fits-all panacea. There are a lot of key decision points that can get missed when you adopt the cloud, so let’s go over a few of those things to consider before pulling the trigger on your cloud adoption plan.
Know WHY
This is the million-dollar question (sometimes quite literally). Before adopting cloud technologies, it’s important to identify what problems the cloud can solve and that you ensure you’re focusing on those goals. By doing this, you’ll save time and money by creating a cloud adoption strategy that works best for your organization and use case.
If you’re not sure if the cloud will solve your problems, discuss this in depth with a trusted IT partner like us.
Understanding pricing
All the cloud vendors have pricing guides, and they’re all just about as easily understood as quantum mechanics. There are permutations upon permutations of variables that go into what any cloud provider might charge you for what you’re trying to do: CPU usage, IPs, data transactional units (DTUs), API requests, bandwidth, storage space, capacity…we could go on. It’s nearly impossible to accurately calculate your overall spend.
A lot of times you really don’t know what the price is going to be until you have your specific application and configuration running in the cloud. Sometimes our clients have been pleasantly pleased, sometimes not so much. This is where doing a proof of concept comes in. Spin it up so you can get a full understanding of the operations and what it’s costing. Identify where you can consolidate or tweak to minimize your costs.
Once you get an optimal configuration, you will have a better idea of your spend going forward (so long as your cloud vendor doesn’t change their pricing). Or you might find out that the cloud isn’t the most cost-effective option, and that’s okay – maybe money is no object and flexibility is.
Identifying your single points of failure
If you move your mission-critical operations to the cloud, whether that be Exchange Online or your database, your internet connection is ultra important. Since it’ll be the only way you can access your information, it’s important to have redundant connections as well as adequate bandwidth.
It’s common to forget considering updating circuits, internet contracts or network backbone. It’s important to make sure you have plenty of resources, like powerful firewalls in a highly available configuration or a secondary internet connection you can fail over to.
Planning for a DNS transition
It typically takes up to 24 hours for full propagation, so plan your maintenance windows accordingly. Prepare ahead of time when you’ll make the DNS changes so that there isn’t a disruption to your services during business hours.
We typically use DNS Made Easy for this, but Amazon Route 53 also works great. They both have A Record failover capabilities to make the transition smooth.
Retooling backup strategy
It’s important to give lots of thought to how you plan on backing up your cloud data. Are you planning on using a specific tool or will it integrate into your current backup solution? Does this change your DR plan?
How do you plan on testing your backups?
Now that you’re paying per-gigabyte storage, have you rethought your retention policy? Are you backing it up to an on-prem device and if so, did you consider the egress fees? Or are you backing it up to the cloud? Do you want all your eggs in the proverbial cloud basket?
And it’s important to keep all of this in mind in the budgeting / costing portion of your planning.
Planning for monitoring of usage
Depending on what you’re using in terms of cloud products, you may need to be monitoring the utilization of your cloud servers. Can you set a schedule to power down your servers during off hours to save you some cost?
It’s important to have a plan to monitor for server sprawl. Depending on your organization, you may have developers who need to spin up machines to do dev/test work, but might abandon the entire instance and spin up a new one when on a new project. It’s up to you to pay attention to server lifecycles and make sure you’re not wasting money running old stuff that isn’t being used.
Having an escape route
What if things are way more expensive than you thought? What if things are way slower than you thought? What if things just don’t work the same? What happens when you’ve exhausted all resources to make your cloud initiative work and it’s time to cut your losses and move back to on-prem?
Do you have a plan to back out or a standby option?
We’ve had many clients come to us asking how they can move back out of the cloud, and it can get messy.
Leveraging the cloud can be a wonderful way to enhance the technology experience at your organization, but it can also go sideways without careful planning. We hope these tips will help you put some more thought into your cloud adoption strategy.