QuestionQ45

VerifiedManage Workloads in a Multicloud Environment

An administrator manages a shared AHV environment where Dev workloads must not exceed 2,000 IOPS per VM during business hours to avoid a noisy-neighbor impact on Prod.

The administrator already uses categories to group VMs (for example, a category such as Env=Dev) and wants a scalable approach that applies consistent throttling to all current and future Dev VMs without configuring each VM individually.

Which approach best meets this requirement?

Explanation

Nutanix Prism Central lets administrators create a Storage Policy that defines QoS attributes—such as a maximum IOPS or throughput limit—and associate that policy with a category instead of with individual VMs. Any VM assigned to the category, whether it exists already or is created later, automatically inherits the policy's QoS settings, and policy settings take precedence over the underlying storage container's defaults. This category-to-policy binding is exactly the scalable, hands-off mechanism needed to cap Dev VM IOPS at 2,000 for all present and future VMs without manual per-VM configuration. Increasing the replication factor changes data redundancy, not IOPS limits; enabling compression reduces storage capacity usage and can affect latency but does not enforce a deterministic IOPS cap; and configuring QoS on a single VM and cloning it only propagates the setting to VMs cloned at that point in time, not to independently created future VMs, so it fails the scalability requirement.

Learn more

Community Discussion

No comments yet. Be the first to start the discussion!