Industry News Release
The original press release can be found here. This version may be edited from the original.
Sunnyvale, Calif., April 4, 2017 – Datrium, the leading provider of Open Converged Infrastructure for private clouds, today announced Data Cloud software, a suite of scalable, secure cloud data management tools, integrated with its award-winning Datrium DVX Open Convergence platform. The company also announced DVX Rackscale systems, providing virtualization administrators turnkey system nodes to speed infrastructure deployment and simplify end-to-end support.
Private clouds are changing significantly in response to increasing use of public clouds. Infrastructure modernization is re-focusing on reducing complexity for virtualization administrators, so traditional systems, including storage arrays, are declining in favor of converged infrastructure approaches. In addition, organizations are relying more heavily on their converged systems for richer data management so administrators have a more cloud-like experience managing infrastructure. Where the goal used to be best-of-breed hardware, now it’s become a race to effortless.
Datrium introduced turnkey DVX Rackscale systems to further simplify webscale or tier 1 private cloud infrastructure deployments. The new Rackscale systems combine DVX Data Nodes with new DVX Compute Nodes, available with either 16 Core Intel® Xeon® E5-2620 v4 2.1GHz processors or 28 Core Intel® Xeon® E5-2680 v4 2.4GHz processors, and pre-configured with up to 768GB RAM, 8 SSDs and DVX Software.
DVX Rackscale configurations can start as small as a single DVX Compute Node and Data Node, and can scale-out to 32 compute nodes per DVX Data Node. Unlike hyperconverged infrastructure, DVX Rackscale systems offer automatic performance isolation, robust security, and accommodate mix-and-match of DVX Compute Nodes with third party compute nodes running DVX Software. The new Rackscale systems are available with end-to-end Datrium Support which provides advanced 24×7 call home and analytics capabilities.
“At Panda Energy, we were looking for a combination of application performance with secondary storage economics, all while reducing deployment and maintenance time,” said Darel Stokes, IT director at Panda Energy. “When we looked at hyperconverged systems, the rip-and-replace of our server infrastructure was going to be a nightmare. With the new rack-scale DVX, we can get the simplicity of turnkey, rack-scale infrastructure while being integrate our existing servers—there is nothing else like it in the market.”
Extending Effortless to the Network
To further simplify rackscale operation and avoid switch configuration where possible, Datrium systems now include innovative Adaptive Pathing software. As networks also move toward commoditization, reducing operator learning curves mean less complexity and better change control. Using Datrium’s end-to-end path awareness, top-of-rack switches no longer need settings such as LACP or MC-LAG for advanced multipath availability and link-aggregated bandwidth. And it needs no configuration settings – it’s always on.
“As enterprises move increasingly in the direction of hybrid cloud environments, we are seeing the evolution of new architectures that are better able to deliver on customer expectations about cloud flexibility, economic and ease of use advantages,” said Eric Burgener, research director Storage, at IDC. “Datrium is unique in the industry for having pioneered a new approach to convergence that fully supports the flexibility benefits of the disaggregated server-storage model, and with Data Cloud is delivering a more cloud-optimized version that will initially support private clouds but lays the foundation for their multi-cloud data management vision.”
Effortless, Integrated Multi-Cloud Data Management
Datrium Data Cloud software builds on the application-centric ease of Datrium’s Open Converged Infrastructure to further collapse the systems administration burden of legacy data management products common to SANs and traditional converged infrastructure. Datrium’s end-to-end deduplication, compression, erasure coding and Blanket Encryption provide the ideal platform for integrating these functions. Data Cloud extends the Datrium platform to integrate functions normally found in third party backup, DR, copy data management and archiving products:
- Dynamic Policy Management for VM-based Data Management
Datrium Protection Groups dynamically bind arbitrary collections of VM, vDisk and file (e.g. templates, ISOs) objects and apply scheduling, retention and replication policies. A Protection Group can include any quantity of objects and even supports overlapping objects with other Protection Groups. Finally, all objects within a Protection Group can be snapped at the same IO moment across hosts for data consistency. To scale, protection group policies are dynamic and inherit all new VMs that fit naming wildcards.
- Copy Data Management for DevOps
In one click or a simple CLI script, VMs can be cloned from existing or snapshot VMs or PGs. Test environments can be quickly cloned from full production builds with minimal capacity and without performance impact, accelerating software development and improving the quality of the result.
- VM-Centric Backup and Restore
Based on Datrium redirect-on-write (RoW) snapshots, backups are granular (vDisk and file level), incorporate dynamic policy management, and include a searchable catalog called a Snapstore. A Datrium system can support hundreds of thousands of VM snapshots. LUN-snap-style browsing for recovery can’t work at this scale, so a Snapstore catalog provides simple, quick searchability.
- Disaster Recovery (DR) and Offsite Archiving
Based on Datrium’s WAN-optimized Elastic Replication, production data can be protected with snapshot–based asynchronous replication capable of providing up to 48 recovery points per day per Protection Group. Replication performance scales elastically with the number of hosts configured. Minimal host overhead is required since only data-reduced snapshot deltas are sent from host flash. Elastic Replication is also fault-tolerant. Unlike HCI-based replication, which stops if a small number of hosts are down, Datrium replication continues even if only a single host remains online. The software also supports 1-to-many, many-to-1, and bidirectional replication topologies.
- Secure Public Cloud Archiving
In a subsequent version of Data Cloud, the company will offer replication to Amazon Web Services (AWS). Based on Elastic Replication, this flexible, secure cloud-based archive will leverage the same on-wire data reduction as well as Blanket Encryption for end-to-end data security between private and public cloud data centers.
“At Neovera, we manage over a petabyte of data and are always looking for better ways to secure, protect and manage that data for our customers. Honestly, we thought we had seen it all in cloud data management until we tried Datrium Data Cloud,” said Scott Weinberg, CEO and founder at Neovera. “Datrium has integrated everything we need in one platform to both run our key performance-intensive workloads as well as handle the data management activities our clients demand. And the copy granularity and policy automation will save us time and improve client uptime.”
“Datrium is moving Open Convergence mainstream by making infrastructure effortless for the VM administrator,” said Brian Biles, CEO of Datrium. “While the hyperconvergence landscape is now more about M&A than innovation, Datrium is rapidly evolving a vastly better, turnkey rackscale private cloud with rich, deeply integrated and scalable cloud data management. And it’s still early in the year.”