Which metrics would you monitor to assess cooling system reliability in a data center?

• A. CPU temperature and energy usage.
• B. Data rate and fiber count.
• C. Return temperature, supply temperature, delta-T, humidity, airflow rates, IT inlet temps, and PUE.
• D. Outside air temperature and wind speed.

Answer: (C)

Monitoring cooling reliability focuses on how well heat is being removed from the IT environment and how the cooling system is performing under load. The best set of metrics includes return temperature and supply temperature to see the heat transfer across the cooling path, delta-T to gauge cooling effectiveness, humidity to protect equipment from moisture-related issues, airflow rates to ensure enough air is circulating through racks, IT inlet temperatures to know what the servers are actually experiencing, and PUE to understand overall energy efficiency and how much of that energy is going to cooling. Together, these provide a comprehensive view of cooling performance and early warnings for potential issues or hotspots. In contrast, CPU temperature and energy usage only reflect individual servers, data rate and fiber count are networking metrics, and outside air temperature and wind speed relate mainly to economizers and do not indicate internal cooling status.