Scaled-Down ROBO Footprint

One of the business challenges with remote offices and branch offices (ROBOs) is that there tends to be a lot of them, so the costs for each branch multiply. The ability to scale an IT infrastructure up or down to meet ROBO requirements is critical to cost-effective branch deployment.

Many hyperconverged vendors have a three-node minimum, but SimpliVity can deliver a highly available branch office with just two nodes, saving one-third of the cost for each branch. A single SimpliVity node at a branch can provide data protection and disaster recovery for an even more cost-effective solution to small branches. In this case, the node replicates its entire workload to the head office or another branch. Both local and replicated backups are available for instant virtual machine (VM) recovery, or file-level restores of Windows files. SimpliVity can scale all the way down to the smallest branch office without compromising data protection and unified VM-centric management of all sites.

In a hyperconverged cluster, a majority of the nodes must survive a failure for the cluster to remain in operation. This means in a three-node minimum cluster, two nodes survive when one fails. However, three physical hyperconverged infrastructure (HCI) nodes may be more costly than a branch office can justify.

To enable a cluster with two physical nodes, the organization still needs a third node, but it doesn’t have to be a physical HCI one. SimpliVity has a small piece of software called the Arbiter. The Arbiter can be a voting member in the cluster without requiring a third physical node at the branch. For a ROBO deployment, the Arbiter runs in the head office and is a voting member for each branch, which means that only one or two physical HCI nodes are required in each branch. In a two-node configuration, a single node failure at the branch will cause VMs to failover to the remaining node. Full VM service will be restored in minutes, automatically. Unique among hyperconverged vendors, SimpliVity also provides hardware RAID to avoid failover when a hard disk or solid-state drive (SSD) fails, as well as redundant power and fans. This removes the necessity of a third node at the branch, saving money as well as space.

Some branches are too small to have two physical nodes, requiring just enough hardware to run a handful of VMs. A single SimpliVity HCI node will provide ample resources, when a second node might be cost prohibitive. With a single HCI node, local high availability will only be within the node. Hard disks, SSDs, fans and power supplies are all redundant. To protect against the whole node failing, the organization should replicate the VMs to another location, either in another branch nearby or to a central office. The organization can replicate the VMs both to another branch and to the head office for additional redundancy.

SimpliVity’s data efficiency allows this replication across low-cost networks. Only new, unique data needs to be replicated, minimizing the required bandwidth. If the branch node fails, then the VMs can be restored at the nearby branch and accessed over the network. Once the branch node is operational again, it is a simple matter to replicate the VMs and then move them back to the branch.

SimpliVity’s approach to data efficiency and built-in data protection enables scaling for small branch deployments without sacrificing availability. A highly available two-node branch deployment immediately cuts costs and provides ongoing savings in power and space. Supporting full data efficiency and protection with a single node is unique among HCI vendors.