A CPU chart, a memory value and a load average answer different questions. Start with the asset, collection timestamp and active workspace. Then compare resource pressure with the workload and recent changes.
Follow the related signals
High load with low CPU can point toward waiting work. Disk capacity is not the same as disk latency. A service can fail while the host remains reachable. E4S places server resources, agent state, service checks and history together so operators can investigate rather than trust a single badge.
Missing data remains unknown coverage. See the resource-signals guide for a review sequence.