Transforming Data Storage: The Leader in Open Convergence for Private Clouds – Datrium, Inc. Sets Eyes on More Markets, Deeper Reach

Infrastructure complexity is at the heart of most data center challenges. While the pace of modern business is accelerating, the weight of aging infrastructure can hold companies back and cost organizations first-mover advantage, market share and even financial viability: Brian Biles

A typical business invests in entry-level accounting software, then adds entry-level or advanced line of business application as needs evolve. At some point, this becomes untenable and you have to evaluate new strategies for the systems that support the business. The goal is to get these siloed applications talking to each other, or even better, running out of one system.

It is at this point that it is important to understand how business systems have evolved and the benefit of cloud computing.

Cloud computing is a better way to ensure you get the applications you need to support your business. Instead of managing your applications yourself, you just sign up and receive service – like you would a utility. This makes it faster, easier and more cost effective to get started.

In view of the above mentioned, we are thrilled to present Datrium, Inc. – dedicated to developing and manufacturing server flash storage systems.

The firm offers Datrium DVX system, a server flash storage system for enterprises; Data Cloud, a cloud-based software for data management; Datrium Blanket Encryption, a software that combines efficient deduplication and compression technology with end-to-end encryption for private clouds; DVX Rackscale systems that offer automatic performance isolation and security, and accommodate mix-and-match of DVX Compute Nodes with third party compute nodes running DVX Software; and DVX system with Flash End-to-End, which takes automatic virtual machine and container performance to high levels offering write throughput per TB while enabling rich data management across snap replicas on flash, disk, or cloud.

The company’s Datrium DVX system includes ESX servers that provide local storage compute and caching using DVX software; and HA appliance that provides simple off-host shared durable capacity. It also offers solutions in the areas of virtual desktop infrastructure, test and development, and server VM consolidation.

Datrium is headquartered in Sunnyvale, California, USA.

Interview Excerpt: Brian Biles, CEO/Co-founder

Brief us about the journey of the company from its inception to date.

Datrium was founded in 2012, began shipping products in 2016. Since then, it has iprise data center. For those organizations who demand large-scale consolidation, high performance and resilience for mission-critical mixed workloads, Datrium DVX delivers a solution--in the simplest possible way.

The pace of modern business is indeed accelerating. Datrium offers IT organizations the opportunity to drive revenue 5x faster, to reduce the time spent managing infrastructure by as much as 95%, and to reduce infrastructure spend by up to 75% or more. There is no other option in the market today for data center modernization than compares with Open Convergence.

Datrium is headquartered in Sunnyvale, California, USA.

Interview Excerpt: Hugo Patterson, Chief Scientist/Co-founder

Brief us about the journey of the company from its inception to date.

Datrium was founded in 2012, began shipping products in 2016. Since then, it has introduced three major product releases evolving the capabilities to meet mission-critical enterprise expectations. Datrium has converged tier 1 HCI and backup, can now scale up to 128 compute nodes and over 1 PB of storage within a single system, and delivers Cloud DR to meet the needs of critical enterprise applications.

What was the motivation to start the company and how did it all come together?

We started with the goal of delivering a dramatically simplified converged infrastructure for the modern enterprise cloud. Melding the experience of VMware and Data Domain, we have produced a new breakthrough architecture that delivers mission-critical converged infrastructure combined with scale-out backup and cloud DR.

How is your company positioned in the current market scenario?

Datrium delivers tier 1 HCI features along with scale-out backup and cloud DR, providing three products in one; a new breed of convergence for hybrid cloud enterprises.

Enterprise customers are struggling to get the performance, ease of use, and cost savings from hyper-converged infrastructure, especially for mission-critical workloads where traditional storage arrays are difficult to implement. In addition, Datrium converges support for these primary workloads together with backup on-prem and to the cloud, eliminating a whole silo of dedicated backup infrastructure.

Datrium delivers a new breed of convergence comprised of dual-tiers: a performance tier and a protection tier which are independently scalable. In addition, Datrium software has been ported to AWS for public cloud DR deployments.

Who are some of your biggest clients? In which vertical do you see the most traction in?

Datrium has clients across many vertical industries including Cloud services, Education, Energy, Financial Services, Government, Healthcare, Media/Entertainment, Manufacturing, Services, and Technology. Major clients include large enterprises like global technology powerhouse Siemens, a global consumer products company, and one of the world’s leading entertainment service companies, as well as popular consumer brands like Oberto Brands and Osprey Packs.

What is the roadmap ahead?

We plan to continue to expand our cloud-based DR capabilities, including cloud-managed

orchestration, as well as support of test/dev and analytics workloads in the cloud.

Testimonials: A Few Of Our Trusted Customers

• Scott Weinberg, CEO, Neovera

“Datrium’s open convergence platform provides us with the kind of elastic performance and capacity that our clients expect with their cloud-based, high-value applications.”

Jim Cunningham, Director of Information Systems, University of Maryland

“We honestly were not expecting to get the kinds of numbers we are seeing with Datrium. The rate of compression and deduplication combined with the low latency is impressive.”

• Joshua Rabe, Systems Architect, North Bend Medical Center

“When I ran a head-to-head comparison, the Datrium system performed critical tasks in a quarter of the time compared to other providers. During maintenance windows, we saw a 5-10x time savings. Tasks that used to take 10-20 minutes on the current SAN now take less than two minutes to complete.”

• Benjamin Craig, EVP and Chief Information Officer, Northrim Bank

“Once I saw the technology, it was like the OxyClean^® of storage, it sold itself. The advantage was instantly clear to me and was exactly what we have been waiting for to finish our virtual desktop infrastructure deployment. It represents an evolutionary step in server-storage convergence by moving the logic, processing, and most of the touch points from the SAN Controllers to each Host Server, while maintaining worry-free data protection on their highly-available Netshelf.”

• Stuart Lewallen, Data Center Team Lead, Sonoma County

“Datrium simplified things so much that my team can focus on other things, whether you’re a small shop or have thousands of machines, this is the way to go.”

• William Hetrick, Technology Specialist, City of Central Point

“Keeping users happy and giving them the best performance is our goal. Datrium has definitely

made it incredibly simple to get us there and keep us there.”

Datrium Solution: Open Convergence

Datrium’s solution does not maintain persistent data in

the server or “compute node.” Instead durable data is written on “data nodes.” Compute nodes use onboard server flash for local reads and writes. They also use compute node CPU resources to perform other services such as compression, local deduplication, erasure coding, encryption, snapshots, and replication.

In contrast to hyperconverged infrastructure solutions available on the market today, customer data is never split across server nodes. That is, data from all VMs on a compute node remains local on that server, removing the network overhead and noisy neighbour problems typical of hyperconverged deployments. 

Datrium uses server-based SSD storage/flash as a fast access layer for data accessed by the VMs on that compute node. As data is written, it’s compressed and locally deduplicated before being written to an object-based, persistent data node.

A data node is a relatively light weight dual controller/HA storage appliance with 12 high capacity disk drives. Data node storage is global to all compute nodes running Datrium software. Besides acting as a permanent repository for data written by the cluster of compute nodes, it also performs global deduplication of data across all hosts.

Open Convergence is easily scalable – if you need more IO performance just add more compute nodes; if you need more capacity just add more data nodes. And if a compute node fails, it doesn’t impact IO or data access for any other host.

Datrium originally came out supporting VMware virtualization, and has since announced support for RedHat Enterprise Linux and CentOS Linux with either KVM or containers.

Recent Enhancements

In July last year, Datrium released new support for RedHat and KVM virtualization alongside VMware vSphere. They also added Docker persistent volume support to Datrium. Now you can have mixed hypervisors (KVM, VMware), and Docker container environments, all accessing the same persistent storage.

KVM offered an opportunity to grow the user base and support Red Hat enterprise Linux accounts. Also, much of the public cloud is KVM based, which provides a great way to support future Datrium storage services in public cloud environments.

To reduce storage management time, Datrium has incorporated sophisticated search capabilities and naming convention options for VM and container policy management. Customers can define a naming convention for VMs, vDisks, and container volumes and use these to define group storage policies, which will then apply to any item that matches the naming convention. This is a lot easier than having to do management at a container level with 100s, 1000s to 10,000s distinct IDs. Each Datrium system includes a backup class catalog capable of maintaining up to 1.2 million VM or container snapshot backups.

The Brains Behind The Operation

Hugo Patterson: Hugo is a co-founder and the Chief Scientist of the company. Prior to Datrium, Hugo served as CTO of the EMC Backup Recovery Systems Division, and the Chief Architect and CTO of Data Domain, where he built the first deduplication storage system. Prior to that Mr Patterson was the engineering lead at NetApp, developing SnapVault, the first snap-and-replicate disk-based backup product. Hugo has a PhD from Carnegie Mellon.

When Hugo left EMC/Data Domain, he was joined by one of the original founders of Data Domain, Brian Biles, and another CTO of Data Domain, Sazzala Reddy. The threesome was subsequently introduced to two early principal engineers from VMware by a mutual friend and early Datrium investor, Diane Greene.