Fifth step in cloud migration – move and build

This is the fifth in a multi-part series. Follow the series here.

Once all the security pieces are in place, cloud deployments can begin. In nearly all cases, we’ve seen customers migrating gradually. They do a proof-of-concept, then migrate their first workload. If they’ve followed the steps in our earlier blogs, this first migration is a success, it is secure, and it performs as expected.

Companies then begin more significant migrations. Using the framework reference architecture they’re already created for the first migration, virtually any VM that’s running on-premises is a candidate for deployment in the cloud. To oversimplify things, the cloud can be thought of as a way to run VMs on someone else’s hardware.

While that overly-simplistic view can help convince skeptics, the cloud in reality is much more than this. Companies can – and often do – deploy entire datacenters in the cloud. Companies deploying mission-critical applications in the cloud frequently turn to products like the Barracuda NextGen Firewall to provide secure remote access, as most cloud providers’ built-in offerings are limited.

Companies also look to write applications specifically for their cloud deployment. These are “born-in-the-cloud” applications that take full advantage of all the extensions and features already built into many cloud platforms. Companies also look to deeply integrated features like artificial intelligence and big data analytics to develop new applications which simply weren’t possible in a completely on-premises world.

But few companies can completely eliminate on-site infrastructure. From a small kernel that provides some level of authentication to high-sensitive data centers or applications that aren’t candidates for cloud deployment, most companies find they still have some level of infrastructure on-premises. Others find that different cloud platforms are more suitable for specific applications; when we ran a survey earlier in the year on cloud adoption, over 60% of the companies we surveyed had at least two different public cloud deployments.

IT organizations need to be mindful that the products they use to manage these deployments – especially those providing security or continuity – can themselves get complicated if different vendors are chosen for different deployments. Vendors who offer products that work across hybrid solutions, and who can support multiple cloud vendors (and there are differences, sometimes significant) can help eliminate the burden that these complex deployments might place on IT.