Nutanix Servers

What hyperconvergence means, briefly

A classic server room has three floors: servers, a storage array, and a storage network between them. Every floor has to be bought, configured and maintained, and growth means adding pieces that do not always get along.

A hyperconverged node folds those floors into one box. The disks sit inside the servers themselves, and software gathers them into a shared storage pool. The separate array and the fabric behind it simply disappear from the diagram.

What that gives you in practice

• Growth by the node. Run short of capacity, add another node, and the cluster expands without downtime.
• One management pane. Virtual machines, storage and networking are configured from a single console rather than three separate interfaces.
• Predictable failure behaviour. A failed node hands its machines to its neighbours automatically, and data lives in several copies.

When it fits and when it does not

Hyperconvergence wins where there are many virtual machines and VDI seats, where infrastructure grows gradually, and where there is no dedicated storage team.

It loses on narrow workloads built around one large database that wants the whole of a single machine, and where a classic storage array is already paid for and earning its keep.

Where sizing starts

Three numbers: how many virtual machines are planned, how many terabytes of usable data, and how many users will work at the same time. From there comes the node count and the series, G8 or G9.

Send us those figures and we will propose a cluster configuration, explaining why that particular node count.