A cloud engineer notes near-capacity CPU and memory across 100 servers with varied usage patterns. Which action will optimize compute usage?

A cloud engineer notes near-capacity CPU and memory across 100 servers with varied usage patterns. Which action will optimize compute usage?

• A. Add additional CPUs and RAM to the host that is serving the cloud.
• B. Adjust the cloud workload by migrating resource-intensive applications to different hosts.
• C. Add additional hosts to the environment using the cloud management tool.
• D. Enable automatic scaling in the cloud management tool.

Answer: (B)

When compute resources are running hot across many servers with uneven usage, the best move is to rebalance workloads by moving the most demanding applications to hosts with available headroom. This targets the root issue: some servers are overloaded while others have spare capacity. By migrating resource-intensive applications, you distribute the load more evenly, reduce contention for CPU and memory, and improve overall utilization without buying more hardware. It leverages the cloud environment’s ability to relocate workloads while the system is running, keeping performance steady and often avoiding the need for immediate capacity expansion. Adding more CPUs and RAM to a single host won’t fix the broader imbalance and can simply shift bottlenecks elsewhere. Bringing in additional hosts increases total capacity but doesn’t address how workloads are placed, potentially delaying improvements in utilization and adding cost. Turning on automatic scaling helps when there’s actual, extensible demand and available capacity to scale into, but with near-capacity resources across the fleet, there may be little headroom to scale out quickly, and it won’t rebalance existing workloads by itself.

When compute resources are running hot across many servers with uneven usage, the best move is to rebalance workloads by moving the most demanding applications to hosts with available headroom. This targets the root issue: some servers are overloaded while others have spare capacity. By migrating resource-intensive applications, you distribute the load more evenly, reduce contention for CPU and memory, and improve overall utilization without buying more hardware. It leverages the cloud environment’s ability to relocate workloads while the system is running, keeping performance steady and often avoiding the need for immediate capacity expansion.

Adding more CPUs and RAM to a single host won’t fix the broader imbalance and can simply shift bottlenecks elsewhere. Bringing in additional hosts increases total capacity but doesn’t address how workloads are placed, potentially delaying improvements in utilization and adding cost. Turning on automatic scaling helps when there’s actual, extensible demand and available capacity to scale into, but with near-capacity resources across the fleet, there may be little headroom to scale out quickly, and it won’t rebalance existing workloads by itself.