CPU, memory, and network utilisation for a server over a 24-hour window on shared axes, so the midday CPU spike can be lined up against memory pressure and network load.