Traditional hyperconverged infrastructure runs compute and storage on identical nodes. This is easy to buy and manage, but it ties the two resources together. An application that needs more capacity forces the customer to add nodes with processors and memory it does not need, and a compute-heavy workload may force the purchase of storage it will never use. Mixed workloads suffer further, because storage services running inside the same nodes compete for CPU with the applications.

Disaggregated HCI separates the resources while keeping the operational simplicity of a hyperconverged platform. Compute nodes and storage arrays scale on their own. Storage processing occurs on the array, rather than consuming server cores, so application performance is more predictable, and the ratio of compute to capacity can be tuned to each workload. This also helps mixed environments in which different applications have different profiles.

Option A describes backup management, which is unrelated to the limitation. Option B refers to single-vendor integration, and no platform ensures seamless compatibility with all workloads. Option C describes a conventional converged or hyperconverged consolidation, which repeats the limitation instead of resolving it. Option D correctly identifies independent scaling of compute and storage, which improves performance for mixed workloads.

Study Guide reference: HPE Storage Fundamentals (HPE0-J81), disaggregated HCI and HPE Alletra dHCI; comparison with traditional HCI.