This guest post is by Saleem Muhammad, Director of Product Management and Marketing at Dell Technologies. We thank Dell Technologies for being a sponsor.
An ever-increasing number of organizations are taking software-centric approaches to IT to better respond to fast-changing business requirements. Such organizations frequently leverage VMware-based software-defined approaches leveraging vSphere and vSAN managed through vCenter.
If you’re an administrator of such environments, you expect your physical infrastructure to be as agile and flexible as your virtual infrastructure. In many environments, however, physical network infrastructure is either run in a silo or is an afterthought, which limits overall IT agility.
The premise of Dell Technologies’ SmartFabric Services (SFS) is straightforward: If you’re an administrator who knows vCenter, your physical network infrastructure should work in concert with your virtual infrastructure, simplifying the operation to just a few clicks and letting automation do the rest.
SmartFabric Services is designed for VMware-based software-defined infrastructures through tight integration with vCenter. With fabric interconnects powered by SmartFabric Services from Dell Technologies, you can quickly and easily provision scale-out fabrics for compute, storage, and Hyperconverged Infrastructure solutions, and connect them back into the data center.
SFS is designed to do three things:
- Streamline device onboarding of servers, storage, and networking
- Provide dynamic fabric operations that protect against human error
- Provide a uniform operational experience for IT shops of any size
SmartFabric Services uses fabric interconnects (Software-defined Interconnects), restful APIs, and industry-standard protocols to dynamically build leaf-spine fabrics for solutions that can start as small as two switches and grow into an 8-rack design with up to 20 switches in a single pod. Ability to manage multiple pods through a single vCenter further expands the size of infrastructure without compromising on simplicity.
SmartFabric Services Embraces vCenter
Network policies for VMs are dynamically applied to interconnects; that is, administrators simply make changes to VMs through vCenter. SFS ties into vCenter via the OMNI plug-in, which a VMware admin downloads onto the vCenter server. This plug-in creates the connectivity between the fabric interconnects and vCenter.
The upshot is that infrastructure admins are able to configure the network from within the vCenter UI in a few simple steps.
Within vCenter, the administrator can assign uplinks and onboard physical devices. When VMs come up on hosts within the pod, they can be easily joined to the fabric.
Switch lifecycle management is also handled through vCenter. For instance, switch firmware and OS updates can be scheduled and pushed out through vCenter.
Ditch The Nerd Knobs
SmartFabric Services hides network complexity by automating typical operations. The goal is to provide a standardized, predictable environment that simplifies deployment and reduces the chances for human error.
Once the interconnects are racked and cabled, an infrastructure administrator has to perform only a single step. The admin issues a simple configuration command to each switch to enable SFS and define the role for each switch: leaf or spine.
This simple process significantly reduces the number of steps required to bring up the fabric. In addition, once SFS is enabled on the switches, all other configuration commands are disabled to protect against human error.
This specific aspect of SFS may help calm the nerves of networking admins in your organization who are wary of having virtualization or server admins tinkering with the CLI. Network changes can only happen through vCenter, mostly dynamically or through a simplified UI-driven approach.
As mentioned, SFS works via the concept of a pod. For a single rack in a pod, SFS creates a LAG at the first layer of interconnects as needed. For multi-rack pods, SFS dynamically creates a BGP EVPN fabric with VxLAN tunnels to stretch L2 across racks. With ESXi and NSX-T integrations the fabric automatically extends layer 2 or layer 3 connectivity as needed. Border switches enable connectivity to networks outside the fabric for workloads that need to reach other applications or services within the data center or out on the Internet.
SFS also supports essential features such as vMotion across multiple physical racks by extending L2 connectivity across the fabric.
Use Cases
SmartFabric Services is beneficial to organizations of all sizes. For small companies that may not have a deep bench of networking experts, SFS provides the benefits of a fabric without the deployment or operational complexity.
Meanwhile, larger organizations may have multiple sites without a large IT presence at each site, or they may have a large data center with multiple workload pods. By tying SFS into vCenter, infrastructure admins can deploy and manage numerous sites from a central location.
You Don’t Need a Network Certification To Configure A Fabric
The SmartFabric Services approach enables infrastructure admins to provision and deploy compute, storage, and HCI fabrics quickly and consistently, meaning network engineers can concentrate on other tasks, while dev and ops teams can provision, distribute, and run applications on known good infrastructure.
SmartFabric Services democratizes infrastructure automation. Whether you’re a small shop with a minimal footprint or a global business that wants a quick way to deploy infrastructure for a department, a branch, or an edge location, SmartFabric Services takes advantage of the vCenter expertise in your organization.