Datrium Sets Impressive World Record With New Breed of Convergence

Hyperconverged solutions solve many issues in today’s data centers, including ease of deployment, agility and scalability, enabling automation, improved data protection and more. But hyperconvergence also has its pain points, with the biggest issue being performance.
Datrium has set out to solve some of the most important issues that hyperconvergence created, including performance. During our recent RoadCast video series, we stopped by the Datrium headquarters to speak with Andre Leibovici, the company’s VP of Solutions & Alliances, to learn more about what Datrium is doing to solve the problems of hyperconvergence for IT architects and server admins.
Leibovici tells us that Datrium’s core tenet is enabling capacity to be decoupled from performance. In typical hyperconverged solutions, when you add capacity, you also need to add more compute, says Leibovici and when you add compute, you need to add capacity, which can create an imbalance. This is one of the major things Datrium is solving. So how does it work?
In the video, Leibovici explains that having the controllers on the servers enables customers to scale up to 128 servers in a single cluster using local SSD and local NVMe so that reads are local; writes are also local, but they go to what the company calls a data node. So Datrium has a node for persistent data that also acts like secondary storage or a backup entity, says Leibovici. In many cases this also solves the organizations’ backup issues, including not having to rely on third-party backup solutions.
With most hyperconverged solutions, organizations are limited by the footprint of the server and how much capacity can be added; but Datrium allows them to grow their capacity as needed. What’s more, Datrium allows organizations to bring in their own existing hardware investments as part of the solution, says Leibovici. All you need is the Datrium software running as part of the hypervisor.
Leibovici also discusses how organizations can move to Datrium and what they need from an IT admin perspective, including how to incorporate existing solutions and hardware.
Finally, Leibovici touches on an exciting announcement on Datrium’s new performance world record conducted by a reputable third-party. In partnership with Dell and, Datrium was able to have its impressive benchmarks validated, showing the solution was able to achieve 8,000 VMs on a single Datrium platform that was running 60 servers and 10 data nodes.
This is truly an impressive result when you consider that the biggest SAN IOmark benchmark was able to scale only up to 1,600, and the highest hyperconverged solution was able to achieve just 800 IOmark VMs. With this announcement, Datrium is able to demonstrate that its solution is not only capable of unprecedented scale, but it’s also the fastest, low-latency solution that is able to keep its data services on, including deduplication, compression and erasure coding.
Watch the video to learn more about Datrium’s mind-blowing world record and check out Leibovici’s blog post on the topic.

Beyond Hyperconvergence

In a separate video we spoke with Leibovici, who was the architect for VMware Horizon, worked for Nutanix and has written many articles, whitepapers and books around his work, to discuss Datrium’s open convergence platform versus hyperconvergence.
Watch the complete video to learn about the major pros and cons of hyperconvergence and where open convergence fits in to solve some of the major problems of hyperconvergence. For more information on Datrium’s open convergence and to get a free demo, visit