“I’m able to do what I did before more efficiently in less time.”
–Health Research Organization
This medical institution once was a typical example of an organization standardized on a legacy virtualization platform to support their various research teams. They had hundreds of virtual machines in their research arm, running a wide variety of workloads, from custom applications to typical enterprise database products. And though the environment was stable, they always felt like something was missing.
What could be missing?
For starters, they had compute virtualization, but were still stuck with legacy storage and networking. After all, the costs of moving to software defined networking and storage were prohibitive, the complexity was worrying, and the benefits weren’t exactly clear. <Read our guide to Software-Defined Private Cloud>
To that point, they were still stuck managing their infrastructure manually. They had some scripts of course, but they needed a storage administrator for the storage, a network administrator for the networking, someone to patch and update the servers, and so on. These requirements, though typical, drove up operational costs and often caused delays. Provisioning storage for a new workload took time, and so on.
Adding new capabilities wasn’t straightforward. For example, their development teams might have been interested in using containers, but adding a container cluster to a legacy virtualization platform isn’t exactly straightforward.
Licensing and support were also okay, but nothing spectacular. The licensing scheme had the usual complexity, support contracts cost money, and if they added additional capabilities, the costs of licensing and support would increase. Nothing unusual, but not very satisfying.
And finally, updates were a nuisance. Every time a major version update emerged, Goran and his team would have days to weeks of work.
All of these problems were real—and frustrating—but they were all part of day-to-day operations and though not ideal, the organization was comfortable. The real inflection point came when Goran started exploring alternatives to legacy virtualization. He realized that his platform, though mature and well-known, was hardly a leading-edge approach and the frustrations he suffered weren’t necessary. He became discouraged by all the limitations and wanted something better.
Goran started to examine a new approach: hyperconverged infrastructure. It was an improvement but had some issues that he couldn’t accept. The biggest issue was the problem of stranded capacity.
Hyperconverged infrastructure consists of servers running software that creates clusters of resources. That’s a simple approach and it works well when workloads scale smoothly, with compute requirements and storage requirements both growing in sync. In many cases, workloads don’t grow in sync— a workload might be compute-intensive but with little storage requirement, or the opposite can be true. That leads to a problem with hyperconverged infrastructures called stranded capacity. As hyperconverged infrastructures scale using servers, it’s common to be forced into buying more compute or storage capacity than needed. Over time, organizations end up with infrastructures at 10% compute utilization and 90% storage utilization, or vice versa. Stranded capacity drives up costs.
Another issue with hyperconverged infrastructures is efficiency. All hyperconverged infrastructures use a clustered approach for data protection. That reduces storage utilization efficiency because data is stored on two or more nodes in the cluster. Data redundancy is always a good idea, but the approach used by hyperconverged infrastructure is unnecessarily inefficient.
Hyperconverged was better, but not right. Goran needed something that was an improvement on hyperconverged technologies.
The company’s IT Director was introduced to Cloudistics. Though we were a young startup, he saw potential in our approach to IT, in that Cloudistics goes beyond the limits of legacy virtualization and hyperconverged infrastructures. It’s a full-fledged private cloud that offers a premium experience for organizations that want to escape the status quo.
- Is software defined, compute, storage, and networking virtualization (hypervisor) are built-in without additional costs or licenses.
- Comes with an Application Marketplace full of operating system and application templates ready for immediate use.
- Automates hardware management, including expanding compute, storage, or networking.
- Automates application delivery, simply and quickly orchestrating resources as needed.
- Offers self-service provisioning to make Agile development easy.
- Supports virtual machines and containers out of the box.
- Simplifies IT to the point that an IT generalist can monitor, manage, and maintain the cloud.
- Includes integrated security, data protection, and analytics
- Comes with a premium support package that bundles white glove deployment, instant access to support engineers, one button log retrieval, and managed patch and update.
All of these are great features and Goran loved them, but he wondered, does Cloudistics address the issues he had with legacy virtualization and hyperconverged infrastructures?
Yes. It did (and still does).
Unlike legacy virtualization, everything was virtualized. He didn’t have to consider the cost of adding software defined networking, it was included, as well as the hypervisor which was a breath of fresh air compared to the complexity of licensing and options of his prior hypervisor vendor. Cloudistics didn’t have hidden costs or a dozen piecemeal options depending on the functions he wanted. Cloudistics doesn’t require specialists, doesn’t involve manual intervention, and didn’t force him into add-on support contracts or licensing.
Unlike hyperconverged infrastructure, Cloudistics scales hardware granularly. He could add compute independently from storage, letting him tailor his physical infrastructure to match the needs of his users. And it also didn’t use clustering for data protection. Integrated RAID and snapshots served his needs well.
In short, with Cloudistics, Goran solved his problems. Today he has over 200 virtual machines running a wide range of research workloads, and he’s influencing the business IT team to consider Cloudistics when they are ready to move beyond legacy IT.
This story is typical for customers who choose Cloudistics. We offer a new and simplified way forward, an all-inclusive private cloud platform that disrupts the costs and complexities of other approaches.