PowerFlex FTW

Tales From the Field – May ‘21

“The Online Retailer”

PowerFlex is the most unique solution in the Dell Technologies catalog. Anyone who’s seen companies transform and modernize their legacy infrastructure with PowerFlex can attest to the array of advantages this product provides.

Most anyone familiar with PowerFlex is also aware of its 5 superpowers:

  • Wide distribution of data
  • Self-healing architecture
  • Built-in multi-pathing
  • Shared storage across all clusters
  • Multi-tiered: storage-only, compute-only and HC nodes

Did you also know there’s a 6th superpower that rarely gets discussed? It’s datacenter consolidation, and it’s an area we’re going to discuss today.

Take for example a large retailer in the Midwest. This company had a wide array of Dell, legacy EMC and competitor storage platforms. They also had a mixture of vSAN Ready Nodes, Nutanix clusters and even a few pockets of VxRail clusters for database workloads.

When we originally talked to them their issues were significant but common with growing customers – their environment was far too complicated, they experienced wild variations in the performance on their disparate workloads and they’d been forced to create silos of technology experts to manage their datacenter.

The sales team identified these issues right out of the gate, and along came our PowerFlex conversation.

Dell was able to show them the simplicity of one technology running every workload. Not only were boot storms a blast from the past on their VDI systems with PowerFlex, but their expected database performance was off the charts with predictable performance for every job. Other workloads didn’t necessarily require the speeds and availability of a new PowerFlex solution, but like the old saying goes, “If you speed up storage, you speed up everything!”. This has never been more obvious than with a PowerFlex solution.

While somewhat skeptical of our holistic proposal, this specific customer decided to take a phased approach to their PowerFlex deployment. They started with a small PowerFlex rack build: base cabinet, access/aggregation/management switching, 3 node management pod and 7 nodes in their two-layer introductory build. The 3 compute-only nodes allowed them to build a greenfield vSphere environment where they started with domain controllers, monitoring servers and other base infrastructure VMs.

Once they were able to test base workloads they were hooked! They very quickly migrated their vSAN Ready-Nodes to compute-only vSphere hosts, with their original 4-node storage pool augmented with a handful of additional storage-only nodes. Their performance was off the charts and their users were very pleased with the results.

This was quickly followed by a migration of Oracle RAC systems to a pair of bare-metal compute-only nodes utilizing back-end PowerFlex volumes. Shortly thereafter the rest of their general workload systems and eventually their SQL farm was migrated to their PowerFlex rack deployment.

We also learned a new term from this staggered deployment: “PowerFlex envy”.

This happens when a subset of customer workloads get migrated to PowerFlex, leaving legacy infrastructure in place for other departments and users. Our PowerFlex customer base is generally ecstatic with the performance PowerFlex provides which leads to this “PowerFlex envy”. Everyone wants on the bandwagon once they see what PowerFlex can provide!

PowerFlex

Over time this specific customer migrated all their disparate silos of workloads onto this specific PowerFlex rack solution. We were able to show them incremental costs for compute-only and storage-only nodes so they could plan for those migrations over time. That approach was very well received by those paying for the solution and made for full transparency for future growth. The days of capacity planning and extrapolating out for 3 or 5 years of growth were gone. With PowerFlex this customer now had a “Just in Time” (JiT) cost for growth as needed, allowing for full transparency across all business units.

Before

After

In the end, this customer had one consolidated datacenter solution in place for all their workloads. They had one lifecycle manager, one HTML5 administration GUI and a system that could expand to as large as their business could dream. All in all, that consolidation effort made all the difference to them and the reason they’re a very pleased PowerFlex customer today.

Share the Post:

Related Posts

%d bloggers like this: