Challgenges Migrating to Hybrid Cloud

Enterprise datacenters that have been around for any amount of time can be very diverse and complicated. Datacenter architectures continue to morph and change adding to the complexity of maintaining applications, storage and delivering new services. Some organizations decide to start fresh in the public cloud to avoid some of these challenges only to find out it's not as easy as it seems. Below are my top challenges when migrating or extending your data center to the public cloud.

1) Refactoring Applications  

I would put refactoring applications as the most enticing option. Applications that are containerized that are built to have the ability to expand in high demand, resilient against availability zone failures and application that consume “just enough" infrastructure seems perfect.

For most customers this has not been a reality. Many application owners will have a hard time tracking all the various dependencies that make their application work end to end. This process tends to lead to cost overruns and failure to achieve spend agreements with the preferred hyper scaler. Many organizations will not have the necessary staff to carry this type of transformation out at scale. Also, many applications won't see the benefit to the business by changing their architecture. At the very least refactoring applications is a question of what is the long-term benefit? We see a lot of customers opting to build a new in the cloud and maintain their existing applications until they age out.

2) Native Instances

All of the major cloud providers offer some way to quickly migrate your servers/virtual machines to their native cloud format. Cloud providers typically limit performance at the instance level and at the storage layer. Typically, customers have not had to deal with these constraints as the servers /virtual machines had full access to the resources that they were provisioned. Two overcome any performance constraints, either the cloud instance level had to be increased and maybe even the provided storage would have to go to a higher performance level. In some cases, you would see storage administrators adding more capacity to get around performance issues. Both techniques would lead to overprovisioning and additional management overhead. These might seem OK doing ones or twos but then it also becomes tribal knowledge for the administrator to keep track of. At scale these “make it work” configurations can be very costly.

Also Cloud A might be the one you want today but how do you get to Cloud B which could be a service provider down the road.

3) Day 2 Operations

The application being moved is only one piece of the puzzle. Organizations have spent countless hours providing support to their end users by automating, protecting, upgrading and delivering remote access to ensure business continuity. Domain expertise from both the private data center and the public cloud provider can be a big jump for one individual that has a lot on their plate. Just like the early days of compute virtualization multiple silos popped up between the storage, networking and virtualization teams. Hyperconverged infrastructure helped a lot to break down those silos but now we have the cloud team being introduced as a new one.

If you’re in a management position I'm sure you can contest to that the hardest problems you have on a daily basis are not technical but merely managing a wide variety of diverse people. Constant well-informed communication seems easy but easier said than done. With moving to the public cloud and the need to have different ways to do networking and providing protection for your applications we'll make change management and even more arduous task.

4) Networking

While networking could have fitted into a couple of the above topics I feel that it needs its own bullet point. When doing even a simple migration to the public cloud the dependencies on an application to have connectivity to existing services can be a lot of work to get right. Most applications were built before best practices of using DNS names instead of hard coded IP address. If you are forced to change the application IP address, it can have a domino effect. This is vastly more important for customers using the cloud provider as their Disaster Recovery location. Such issues tend to be fully unmasked at the time of failover.

In a lot of cases the private data center isn't going to be totally going away and so there needs to be communication between the hyper scaler and the private data center. Even in the case of a migration there is going to be an extended period where two-way communications to the cloud workloads and the existing workloads in the private data center needs to be working. Support around existing firewall usage and support needs to be addressed. New security paradigms from the cloud side like security groups or ensuring accessing cloud services are not directly coming over the web need to be planned as well.

5) Time

How fast that you need to migrate out of a data center will be a challenge and will help determine what path you take as you start to move workloads. While this may be given in any scenario you want to ensure that walking back any plan is not going to be costly if you made some decisions in haste. Having flexibility in how you migrate your applications will save you time and money.

Nutanix Cloud Clusters(NC2) is uniquely positioned to address a lot of these challenges by providing consistent operations for both the private and public cloud. Nutanix’s ability to deploy onto hyperscaler bare-metal allows Nutanix AOS to be the common denominator for both private and public clouds. Businesses that consume NC2 we'll have an easy way to migrate applications and still have access to all the great cloud services at their disposal. As applications run on NC2 they will be able to connect both to on Prem and cloud services while keeping their original IP address. When application teams have time, they can continue refactoring the applications that make sense to the business versus trying to do them all. once customers realize they can save money versus native instances due to the virtualization and performance benefits of Nutanix some customers are opting to keep their high-performance workloads on Nutanix versus native services. One such service has been the Nutanix Database Service. Guaranteed performance can be very costly in the cloud and this is just another area that NC2 excels in.

Please consider forwarding on some of these technical videos to your team members that highlight Nutanix’s ability to address a variety of public cloud needs:

Quick Videos

·        How to Lift and Shift Into Azure

·        Secure NC2 on AWS with Custom Security Groups

·        Bursting 2,000 Citrix Desktops to AWS in Under 2 Hours

·        Multicloud Migration: Move Apps to Any Cloud at Any Time

Tech Notes

·        Nutanix Cloud Clusters on AWS - Solution Tech Note

·       Nutanix Cloud Clusters on Azure - Solution Tech Note

·        Disaster Recovery with NC2 on AWS - Nutanix Validated Design