Datrium review 2019

Datrium review 2019

About Datrium

With our Datrium review 2019 we will see that it is a relatively new company that started in 2012. They were started by Data Domain and VMware key staff. Datrium had a mission to bring to market a solution that delivers on the promises of HCI without its inherent compromises. Datrium’s solution is called DVX, and it is a “System for Rack scale Convergence”. You can kind of compare this to a few HCI systems out there and one would be Dell EMC’s VxRack SDDC solution. “The latter would be the better one.

Datrium DVX

In our Datrium review, they make the DVX of two separate hardware components. 1. DVX Compute Nodes (1-128 nodes), and 2. DVX Data Nodes (1-10 data nodes) as of Oct 2019. These two are combined by Datrium’s own software layer. The compute nodes are x86 like all the other SDS HCI compute nodes out there. The compute nodes run VMware ESXi, flash drives and NVMe up to 16TB (RAW) per host and is used for local read-cache. Customers can leverage their existing servers as part of the solution, allowing for both new or current servers, rack or blade. Another useful feature is “Peer Cache mode”, it allows VMs on an affected host to use flash on a neighboring host until it resolves the issues. The all-flash local storage tier combines up to 18 Million IOPS (4kb) and 256 GB/s random-write throughout. This is what they tested and the highest ever audited results in history with IOmark. This is according to Datrium as you can get better throughput with a SAN like the PowerMax but this is in a different scenario.

Datrium HyperDriver


Datrium’s HyperDriver enables the host to use dedicated CPU power to control I/Os to and from the external Data Node. as a result, this enables it to have advanced Data Services such as Compression, Deduplication, Encryption, Erasure Coding, Snapshots, and Replication. To achieve this, Datrium DVX nodes reserve a minimum of two CPU cores to exclusively perform data-related processing. If the hosts have more cores, the amount of CPU reserved for data processing can be increased up to 20% CPU capacity of the node in “Fast Mode” and up to 40% in “Insane Mode”. The naming sounds like a marketing scheme 🙂
You can think of it as the quality of service for VM’s using CPU and storage which is noteworthy. The customer also has the option to use SDS on the host or to use split-provisioned storage. This is cool as you do not have this option with solutions like vSAN, Nutanix or SimpliVity.

Click here for technical detail about their compute node

Datirum’s Compute Nodes are stateless. I am saying it can go down and it keeps high availability. This is versus the 2-3 server failure in most other HCI solutions.

Click here for technical detail about their data node

New DRaaS

As of Aug 2019 Datrium’s DRaaS with VMware Cloud on AWS. Built on Datrium’s orchestrator, ControlShift, this SaaS solution provides the world’s first production support of fully automated fail-over to the VMware Cloud on AWS based on snapshots held in AWS S3 according to them. Datrium DRaaS offers comprehensive disaster recovery (DR) and cloud backup for data protection of on-premises Datrium DVX systems, with Datrium providing a single point of ordering, billing, and support. Thus eliminating the need to manage AWS and VMware Cloud on AWS accounts directly. Datrium DRaaS enables organizations to make the cloud their DR data center at less the cost of a second data center, delivering cost savings. They can ensure companies using Datrium DRaaS that their DR plan will provide fast fail back, granularly recovering loads of VMs after a sudden disaster or a rolling disaster such as a virus attack, and enjoy reduced complexity.

Cloud DVX has scaled customer capacity from 30TB to 1.15PB in the cloud.

ControlShift, a disaster recovery orchestration functionality that can provide a five-minute recovery point objective (RPO) with zero recovery time objective (RTO), offering fail over from cloud or on-premise DVX deployments to Cloud DVX or VMware on the Amazon cloud. They plan Microsoft Azure support for 2020.

ControlShift also offers continuous testing of DR infrastructure.

Conclusion

With RedHat and Pivotal, Datrium allows customers to embrace Kubernetes and containers across multiple virtualization platforms. As a result, they have a pretty cool HCI solution for specific work cases with stable and tested features, and they have a unique solution.

By Cobus Smit – 2 Oct 2019

Spread the HCI Information

This Post Has One Comment

Leave a Reply