"Koehler said VMware has since revisited its testing regime because, "While useful, synthetic tests do not reflect the characteristics of real world workloads and the behavior of the storage system." His team therefore "pored over telemetry data gathered from thousands of vSAN clusters running all types of production workloads" and found vSAN clusters use "much less RAM than expected" and "may use fewer CPU resources than expected.""
"VMware's latest guidance for hardware used in vSANs therefore has lower minimum specs for "ReadyNodes" - the storage-centric servers used in virtual storage arrays - a welcome change at a time when increasing memory prices are making servers more expensive. The guidance could also mean vSAN users can get away with using fewer ReadyNodes. Koehler also pointed to the potential for vSAN designs based on VMware's new recommendations to consume less energy, and therefore reduce cooling requirements."
VMware acknowledged that prior vSAN hardware guidance was based on synthetic testing designed to measure extreme performance scenarios. Telemetry from thousands of production vSAN clusters showed real workloads consume much less RAM and may use fewer CPU resources than expected. As a result, VMware lowered minimum ReadyNode specifications for vSAN deployments. The reduced specs can lower server costs amid rising memory prices, potentially allow fewer ReadyNodes per cluster, cut energy and cooling needs, and improve rack placement and network efficiency. The changes may make Cloud Foundation adoption more financially attractive despite recent licensing shifts.
Read at Theregister
Unable to calculate read time
Collection
[
|
...
]