Hyper-converged architectures are solutions that leverage virtualization to support a software stack that aggregates storage and networking across compute nodes, converging these resources into a single stack. These solutions and their potential have been the focus of a lot of hype, but now some are demonstrating a level of maturity that should lead to wider adoption in the data center. SimpliVity with the latest release of its OmniStack 3.0 is an example of a vendor demonstrating that maturity.
One of the first challenges that hyper-converged vendors face is defining what hyper-convergence actually is. Many vendors claim to have converged and hyper-converged architectures, but these are often independent hardware solutions that are pre-integrated prior to arriving on the customer’s dock. As we described in our article “What are Converged Infrastructures?” these are “hardware converged” architectures, not hyper-converged architectures. Some hyper-converged vendors provide a turnkey solution (compute, networking, storage and software) while others provide just the software. In either case the focus is, as it should be, on the software and that is at the heart of the SimpliVity announcement.
Improved Data Protection
SimpliVity already has built-in data protection that provides backup intervals of as little as 10 minutes. But it was for full VM data protection and full VM recovery. To restore a single file required either restoring the full VM or leveraging a third party solution from Kroll OnTrack. With OmniStack 3.0, SimpliVity adds file-level restore capabilities. Essentially, the VM is presented as a DVD media that can be browsed and recovered from.
In addition to file-level restore, SimpliVity now provides the ability to create and edit backup policies at scale. Multiple VMs can have their protection levels created or adjusted in bulk. The 3.0 release now allows the retention time for existing backups to be edited and the retention time for manual backups to be set upon creation. For example, if the SQL administrator wants to make a quick backup before a major code update, they can perform a manual backup but only have it retained for a few days. The advantage is a more efficient use of backup capacity. Finally, backup policies can be suspended and subsequently resumed so that they don’t interfere with a maintenance window.
The net impact of these changes evolves the SimpliVity data protection capabilities from which could be thought of as an insurance policy to a complete, and fairly robust, integrated backup solution. While a second product might be used for long term data archive, the need for a daily backup application in addition to SimpliVity’s is essentially eliminated.
Enhanced ROBO Support
Another benefit to this new release is SimpliVity’s adoption of a federation topology, ideal for customers with multiple data centers spread out globally. This hub and spoke topology enables nodes to be added to a SimpliVity cluster and have it understood by other clusters in other locations automatically. This allows for the scaling of a data center’s secondary location to be more easily managed remotely. Essentially, all the administrator needs to do is turn on the cube (SimpliVity’s term for a server node) and it will come up in a “waiting to be discovered” mode. Then SimpliVity’s cluster can be used to add the new node to the particular data center it resides in. The alternative is a mesh topology which requires that each site be made aware of the new node and what its role is in the cluster. As the number of sites and nodes increases, this becomes a time consuming process.
As part of this new ROBO support SimpliVity has introduced an appliance specifically targeted at this market. SimpliVity’s OmniCube CN-1200 is a compact system delivered in a 2U form factor. It is designed for smaller deployments and smaller workloads. It offers 8 CPU cores, 50 to 82GB of usable memory, and approximately 2.4TB of effective storage capacity. The OmniCube CN-1200, when deployed as part of SimpliVity’s Unified Protected ROBO solution, provides tremendous agility, simplicity and data mobility.
The 3.0 OmniStack release also improves an administrator’s ability to track resource consumption at a VM level. This allows them to identify VM workloads that are generating performance spikes and to pin-point performance irregularities. Parameters like throughput, IOPS and latency can all be tracked, and VM specific attributes like reads, writes and total I/O can be reported on.
Data protection often seems to be the last step on the way to creating a mature architecture, yet SimpliVity has focused on it from the beginning and with OmniStack 3.0 has extended their capabilities in this area. While the federated topology and performance analytics will benefit all of their customers, the company is clearly positioning itself to meet the demands of the larger data center. Companies like Cisco and Lenovo have established programs that allow the companies to work together. These relationships provides SimpliVity customers with more choice of which x86 system platforms to run the SimpliVity hyper-converged infrastructure and shows further sign of a maturing offering. Presenting a mature hyper-converged offering is going to be key for these larger data centers, and with 3.0 SimpliVity has taken a big step in that direction.