For any organisation, the transition to Cloud or adoption of Cloud is a hot topic that raises many a question and concern that is likely to slow down proceedings. In the earlier days of cloud platform availability, there was a much larger push for migration of servers to the cloud which left consumers scratching their heads and raising larger questions and concerns around the point and benefit of adopting a cloud for this type of service at all. Migrating and running servers hosted in the cloud can still surprise you with a much bigger bill that you might have been dealing with when your servers remained on-premises. That is of course, unless you adequately plan your migration and hosting strategy to utilise the most efficient tools and services that Cloud has to offer.
Fortunately, there are now better tools and methods available to analyse and migrate your servers, plus a much larger range of hosting options for running your virtual machines when they get there. When planned well from the beginning, with a re-aligned mindset that moves away from building on-premises in the cloud, migrating your servers into a cloud platform can actually start to look much more appealing from a management and cost perspective.
This is not to say in any way that migrating servers to cloud will always be a better solution than running them on-premises. You might have reached a point where a hardware uplift is required so cloud is being considered as an alternative, and in any case like this then an assessment needs to be carried out to validate the best option moving forward. This assessment would have to consider Platform as a Service options, however it must also consider Infrastructure as a Service with modern options that won’t be present to you when you plug your estimated server running costs into a cloud calculator you found through a Google search.
Consider the following for your cloud server migration project:
The most costly mistake you can make when migrating servers to cloud is to lift and shift without reassessment. Cloud is not an extension of your on-premises data centre, but if you treat it that way then you can expect some bill shock. If you want to minimise costs when running servers in cloud, you will need to develop an analysis strategy that starts with the on-premises server and ends in continuing analysis long after the server has been migrated and cut over.
In the past when we have built virtual machines on-premises, the usual practice is to allocate it some processing power and then add to it when better performance is required. I have never known a client to reduce processing power when a server has reduced its workload. However, when you start paying for a virtual machine by the hour, you want good analysis tools in place to tell you that after one year of running costs, your server isn’t needing what you have allocated it and what you have been paying for. The terminology here which encompasses this ongoing analysis is virtual machine ‘rightsizing’. A term you are going to get very familiar with if you want to save money on virtual machine cloud spend.
Network to cloud:
When we manage our own data centres, we have a feeling of being in control of our infrastructure and networking but as soon as you start talking cloud that feeling can immediately turn into uncertainty for varying reasons. You are considering sending large quantities of your secure corporate data over a link owned and managed by someone other than your team, which understandably is going to incite roadblocks from relevant internal parties. These roadblocks can be eased and removed in turn with enough knowledge used to create a suitable networking strategy.
We are not talking about a data centre migration here, where there is ease in carrying physical storage from one location to another to plug in and transfer your data (though this is a possibility with some providers). We are talking about sending potentially gigabytes, terabytes, petabytes over a network link that needs to have adequate bandwidth to transfer your data securely. The chances are, your existing internet bandwidth wasn’t built with migrating these amounts of data in mind, nor does it pass your security tests for the class of data travelling over it. You now need to assess whether using a VPN or a direct cloud connection (e.g. Azure ExpressRoute, Google Dedicated Interconnect) will suffice, and this assessment against your personal requirements can be critical as it might dictate which cloud provider is going to be the winner to host your virtual machines.
There is a vast range of cloud migration tools out there to assist your organisation with migrating your servers to cloud. While it is great to have options, it can make the decision making process much more convoluted without some helpful guidance. That is, if your strategy proves that an online migration is technically and financially better for your server migration requirements over an offline server migration. With this proven, you can skip the use of a migration tool in place of your cloud provider’s native offline migration tool.
Each cloud provider offers a native migration tool to get from your on-premises location to their cloud storage. In the short term you will need to know how and if these tools will work with your on-premises infrastructure: VMware and Hyper-V, Windows and Linux version compatibility. Planning a longer term strategy though you can get off on the right foot by putting in place a multi-cloud tool that offers more flexibility with no vendor lock-in. Additionally, when planning your migration strategy, the tool you use can double up as your cloud optimisation tool. Getting this strategic step correct can be key to making sure you don’t overspend in your migration process and virtual machine cloud hosting costs, in the short and long term.
Moving your on-premises servers into a cloud platform is much more of a transformation than a migration. For any kind of transformation if you don’t have a solid strategy in place from the beginning, you can expect a bumpy and costly ride. Not forgetting, this strategy begins well before you can move a server into cloud: for example, have you defined your cloud operating model? Is your cloud governance framework planned and in place? How are your servers going to be protected once migrated? Establishing these models early will help you develop a manageable and financially viable cloud server migration project.
It doesn’t matter where you are in your cloud journey, defining strategies in needed areas will vastly improve your situation. Diaxion have experienced and certified Consultants that can help talk you through your migration strategy for a better outcome with a focus on where you want to be positioned in the long term.