In-Depth

Hyperconvergence for ROBOs and the Datacenter

Remote/branch office management is more important, and complicated, than ever. Here are some survival tips.

• By Mike Matchett
• 12/17/2015

Convergence is a happy word to a lot of busy IT folks working long hours still standing up large complex stacks of infrastructure (despite having virtualized their legacy server sprawl), much less trying to deploy and manage mini-data centers out in tens, hundreds, or even thousands of remote or branch offices (ROBOs).

Most virtualized IT shops need to run lean and mean, and many find it challenging to integrate and operate all the real equipment that goes into the main datacenter: hypervisors, compute clusters, SANs, storage arrays, IP networks, load balancers, WAN optimizers, cloud gateways, backup devices and more. From a logical perspective, when you multiply the number of heterogeneous components by a number of remote locations, the "scale" of IT to manage climbs very fast. If you factor together the number of possible locations and interactions, the challenges of managing at scale can grow non-linearly (i.e., exponentially).

Vendors like VCE, Dell, HP and IBM sell datacenter convergence solutions that get a lot of attention among big IT shops looking for ways to accelerate IT implementations. But racks of disparate equipment, even though shipped as an already configured "unit", can still bring some challenges down the road.

In order to offer an even more "plug and play" solution, some vendors have gone the next step with "hyperconverged" solutions in which all the myriad required IT resources are fully baked into Lego-like bricks of infrastructure. Hyperconverged IT appliances from the likes of SimpliVity, Nutanix, Pivot3 and Scale Computing offer greatly simplified management, operations, and growth that can help deliver consistent, reliable service while offering compelling TCO arguments that include both CAPEX and OPEX savings. "Reference Hyperconvergence" appliances are also available from integration partners of software-defined vendors like Maxta, HP (StoreVirtual), and VMware (EVO with Virtual SAN).

ROBOs Need Edge Convergence
Still, converged branch (e.g., Dell VRTX) or datacenter-style hyperconverged appliances deployed out to thousands of ROBOs can severely challenge a small central IT staff. A full stack/appliance still needs to get deployed to each remote location, often in triplicate for availability, and then needs local physical and data protection beyond what might be practical or cost-effective.

For the IT group trying to be highly efficient, it's still hard to deploy, secure and fix mini datacenters at a large number of remote locations, much less prevent any network connectivity issues. And up until now, traditional convergence solutions deployed at ROBOs didn't further help protect remote company data that might be created and/or need to be immediately available in the remote locations (anymore than non-converged solutions that also remotely replicate or backup).

To be clear, a big ROBO problem is that remote networks are simply not fully reliable and therefore some applications, with their data, still need to be running remotely for availability reasons. Other apps might need to be directly local for security or high performance. While remote desktops might be good enough for a wide swath of mobile or remote users, ROBOs almost by definition require onsite computing. But then how to best address ROBO data storage?

Ways to Meet ROBO Data Challenges
Common approaches to managing ROBO data challenges might include some kind of offline remote backup scheme; or even better, a cloud storage gateway. For example, Connected Data offers a file sharing approach in which each appliance, when deployed to remote locations, automatically joins together with the rest to form a distributed private file storage cloud. This self-forming cloud essentially offers a private Dropbox-style internal service.

Ctera is an interesting converged ROBO cloud gateway solution for file services. Their latest appliance marries local cached NAS file services, integrated backup, and enterprise file sync and share capabilities. This enables a centralized storage IT shop to widely share and protect critical file data across many ROBO locations.

When it comes to high performance ROBO applications that need Tier-1 primary block storage (e.g., anything that needs a high performance database), Riverbed's SteelFusion provides a highly scalable converged ROBO solution. These remote appliances have "hyperconverged" Riverbed's world-class WANO with a hypervisor host (vSphere) and what can best be described as full online remote data "projection to the edge," where all live Tier-1 data at the ROBO is really served out of central enterprise SAN storage. Critical virtual machines and their primary storage can be transparently served out of the data center, but then actually run locally in the branch. In this way, ROBO operations become resilient to most network issues, while eliminating the risk of remote data loss.

Making ROBOs the Smartphones of Distributed IT
Overall, the real challenge for IT is how to deliver virtualized applications in a consistent manner, regardless of whether the user is supported directly by data center, has a cloud-based virtual desktop, works on a mobile device, or sits in a ROBO. In our view, this makes physical infrastructure really a secondary concern; it should be invisible to the end user, and just like our smartphones, deliver applications to us wherever and whenever we want them.

In fact, companies running ROBOs would do well to look for solutions that enable them to deal with their ROBO IT management, remote user needs, and corporate data protection requirements as if those ROBO's can come and go as easily as I lose and replace my smartphone.

The IT infrastructure deployment at a ROBO should be able to spun up or down at will, remotely upgraded and recovered easily, even at another location. To that end, IT really needs to view themselves as measured and valued ultimately by application delivery, not as to how much or how complex or what kind of infrastructure stack they manage day-to-day.