How Do You Configure HA and DRS in Production?

Tobias Whitmore

In your real-world environments, do you prefer fully automated DRS or partially automated mode?

I’m curious how others balance workload mobility and stability when running mission-critical VMs.

Everett Kensington

We usually run fully automated DRS in production, but with a moderate migration threshold. For mission-critical VMs, we also use VM/Host affinity rules to keep certain workloads stable.

HA is configured with admission control enabled to avoid resource overcommitment during failover.

Tobias Whitmore

Everett Kensington We usually run fully automated DRS in production, but with a moderate migration threshold. For mission-critical VMs, we also use VM/Host affinity rules to keep certain workloads stable.

HA is configured with admission control enabled to avoid resource overcommitment during failover.

Jarvis Holbrook

We usually go with partially automated DRS for mission-critical workloads. It gives us more control over migration timing while still benefiting from intelligent placement recommendations. Stability always comes first in production.

Tobias Whitmore

Jarvis Holbrook That makes sense. Partially automated DRS definitely provides better control over migration timing, especially for sensitive workloads. Keeping stability as the top priority in production is always the safest strategy.

S_Ashwood

In our production environment, we use fully automated DRS with properly adjusted migration sensitivity.
Mission-critical VMs need steady performance and automatic load balancing during busy times or host problems, while manual changes just add delay and risk.

We do use anti-affinity and host affinity rules for key workloads to keep things stable.
Overall, full automation gives us better uptime and less manual work without hurting reliability.