As IT organizations look to be more agile in supporting ever-changing business requirements, they’re turning to Kubernetes to speed up their delivery of value. However, with Kubernetes being an open source software, IT departments might have too many options for modernizing with this platform.
Let’s use the fictional monster Frankenstein as an analogy to explore the challenges in building this modern framework.
Imagine that Kubernetes is the heart of Dr. Frankenstein’s monster, a creation built to be a better version of man. Let’s also imagine that private cloud is the entire “Improved Man.” Private cloud is intended to be a better version of the traditional on-premises data center.
While the heart (Kubernetes) is vital to the Improved Man, it’s just one piece of the larger system. And while Dr. Frankenstein’s intent was to stitch together a myriad of appendages to create the Improved Man, what he ended up with was a monster. IT teams need to take heed of Dr. Frankenstein’s story. Their efforts to create a private cloud using Kubernetes may result in a monster.
Dr. Frankenstein proved two things when he created his Improved Man. First, improving on well-designed systems is no easy task, even for a mad genius. And second, even with good intentions we can still create a monster. Disclaimer: We’re not calling engineers and architects who have functioning Kubernetes environments mad geniuses.
We are, however, pointing out that it takes many stitched together software stacks and thoughtful effort to design and empower Kubernetes to support production workloads. And this is to say nothing about the ongoing efforts to maintain these new systems. Many a thoughtful engineer has been lost to the subsequent duty of keeping the monster alive.
To demonstrate the complex nature of integrating the pieces of the Improved Man, let’s look at an important appendage: the brain. After all, it’s important that the Improved Man process and store data as well or better than current man. Dr. Frankenstein thought he could send his trusty minion to collect any old existing brain.
Many data center storage teams trust their OEM vendors in the same manner, but instead they get back an abnormal brain. For decades, they built applications upon the foundation of those traditional SAN and storage vendors, which offered a rich set of features including high availability and incredible performance. The storage team now assumes those same sets of features will be in place when they start building their Kubernetes cluster.
All vendors, with minimal effort, can update their website to claim compatibility as a storage device, but are they really ready? Is their “brain” ready to be host to the most valued component of your company, which is your data? Does it have the required performance, replication and intelligence for backup? Does it know that its interactions must be dynamic and API-driven because development has moved from pets to cattle for the application design? Is it able to restore the entire decoupled application and trust that the intelligence of the application extends to your DR and SLA needs?
This landscape is maturing quickly as features and requirements emerge and change in the middle of our shift to this new agile application delivery methodology. Do you want to spend your valuable time and resources performing lobotomies, which risk the life of your Improved Man?
Improving Data Center Agility
Over the last 20 years, VMware has perfected their version of the Improved Man. Virtualization of the infrastructure more efficiently utilizes compute resources and is a better way to support applications than bare metal. We are, likewise, at another inflection point. Modern applications and app development requires a shift to more agile and portable frameworks.
It’s no surprise that many organizations have tried to tackle this challenge on their own. Often, IT departments never completely get the Improved Man working and end up with a monster. A monster that occupies most of the IT department’s time just to maintain, let alone improve. So that begs the question: Is there a better way? Can we empower the IT organization to develop a modern data center and avoid creating a monster?
Driving Improvements with OpenShift
That’s where OpenShift comes in. Red Hat is the mad genius that has the skills and focus to create a “new man” that works as intended. A data center that IT teams can use and command, allowing them to focus on the value Kubernetes brings to their organization, instead of constantly performing triage.
OpenShift allows IT teams to reserve their evil cackling for insights from data and projects completed on-time and on budget instead of half-hearted chuckles from the limited joy of seeing the monster breath once again after an upgrade. Join CDW’s DevOps team ― grab a pitchfork or torch and storm the data center to eradicate those monsters ― and march forward as a new and agile future.