Going to the Cloud? Fasten your seatbelts first.
So, you’ve taken the plunge. After years of reading about and listening to all the Cloud hype, you’ve finally decided to move your cherished business critical on-premise systems to this mystical place where you have minimal costs with maximum flexibility – what is there to lose?
Given the drive from all CSPs (Cloud Service Providers) to increase their market share, it’s no surprise that they are picking the most interesting and attractive aspects of their services to push to their prospective and existing clients. And there’s no doubt that CSPs can offer some very attractive products and added value compared to traditional on-premise infrastructure – cost, availability and flexibility to name just three!
In this blog post we’re going to look at one of those factors; availability, and how it might be important to look a little closer before making the jump.
Moving the responsibility from the IT department to a cloud vendor
In a traditional on-premise environment it would be the responsibility of the IT department (or supplier) to manage all aspects influencing the availability of the system. In the case of business critical systems, they would be responsible for ensuring that the systems as well as supporting infrastructure were redundant, that there was redundant connectivity from each location to these systems, that the backups were regularly running, and that critical data was stored offsite in case of a disaster. They would be responsible for building, documenting and of course testing all of this to ensure that everything was working as expected. All of this carries significant effort, and of course cost. But all of this cost would disappear once migrated to the cloud – all that’s needed is an Internet connection from each remote location and we’re good, right? Wrong.
Three key factors to ensure security in the Cloud
A common misconception regarding Cloud services is that they are inherently secure and fault tolerant; you put your data in the cloud and it’s available anytime, anywhere, anyhow. That is not necessarily the case. It is true that at almost all levels CSPs will provide redundancy at the virtual machine (VM) level but that is often where the ‘out of the box’ redundancy or fault tolerance ends. For a business critical system you’re going to want much, much more than that. Let’s take a look at some of the key factors you need to consider when moving business critical infrastructure to the Cloud:
As already mentioned, often VM redundancy will be in place as standard; that’s one of the key benefits of VMs over traditional physical hosts. But do verify whether the machines you’re using are covered by this: If you have extremely large or heavily customised VMs they may not have host fail-over options, or of course if you’re using ‘bare metal’ devices then you cannot have inbuilt redundancy at the machine level.
Even if you have VM redundancy, what about the redundancy at the datacentre (DC) level? Do you have options to move your machine to a different part of the DC (fire/water/electricity segregated)? What about at the even higher level? While the DCs of top tier CSPs are extremely resilient, natural disasters or socio-economic effects can still render the whole DC unreachable. Many CSPs offer multi-regional options for redundancy but this is far from ‘out of the box’ and can carry significant cost and effort overhead.
While no one is surprised that ‘free’ cloud storage at the consumer level is not backed up, when it comes to your critical data you’re going to want to make sure it is properly catered for. Naturally CSPs have multiple options available from legacy backup solutions to more modern cloud backup alternatives – but this is all something you’ll need to investigate and consider prior to deployment.
Don’t forget the need to get access to the backed-up data from the locations you need it – especially important when coupled with geographic redundancy as mentioned in the previous point. When dealing with vast quantities of data (GB, TB, and so on) do not underestimate the time it would take to transfer the data from one location to another in the event of a restore being required. With an on-premise setup you may have been able to physically ship terabytes of data from one location to another much faster than an electronic copy, that is not a simple concept when working in the Cloud (although there are some options available).
3. Disaster Recovery (DR)
Setting up the necessary infrastructure and processes to provide resilience is one thing, but how to enact them in the case of a disaster is another. What kind of automation is available, how quickly can the system respond, how low can your RTO get? All of these questions require significant time and effort to determine together with your chosen CSP, and for any systems more complex than single VMs there is no ‘out of the box’ option that you can click and forget about.
Don’t forget about your data! It’s easy to focus on the DR aspects of the processing part of your infrastructure, but the best and most resilient setups will be of no use if your data is not also available in the event of a disaster. As mentioned above, do not underestimate the demands placed on your Cloud setup to keep your data appropriately synchronized to be available in the event of a disaster!
Testing is a crucial part of any infrastructure design, however DR testing is one of the items that is made significantly more complicated in the cloud. Instead of only having to consider your own business, any tests that impact more than your infrastructure in the cloud are going to be next to impossible to coordinate. CSPs are taking measures to try to allow customers to execute tests virtually (such as logical isolation of systems), but that is still not the same as a full DR test to determine how everything works under real world conditions.
Plan ahead to safeguard your business-critical systems
Reading the points above you might jump to the conclusion that moving to the Cloud is not a good idea. That’s not the case at all. The Cloud can bring significant benefits in many areas, the purpose of this post is to ensure that you do not head carefree into a major project impacting critical business processes having been blinded by marketing.
Instead, it’s more important than ever to carefully plan the project and ensure you have the right team and expertise to hand in order to traverse the confusing and potentially misleading minefield of different options, products and price lists. Thankfully, an entire industry has evolved designed specifically to help companies move infrastructure to the cloud and to make sure they’re taking advantage of the latest developments and improvements.
At Datavard we combine decades of SAP and data management expertise with experience from all top tier CSPs to enable our clients to get the best expert advice for any of their ‘move to cloud’ projects. Whether it is deploying complex infrastructure, transforming your existing on-premise environment, or even just leveraging the Cloud to expand your existing infrastructure – we have the tools and the experts to help you achieve your goals! For more information about our methodology, download the “Lift’n’Shift to the Cloud” solution brief.