A server appears to be writing data to a storage system slower than it used to. Which of the following practices would BEST help to validate that there is a problem?

Setting alerts is a reactive measure for notification when a problem occurs, but it does not proactively validate whether a current observation is a problem without a baseline.

Setting thresholds defines limits for performance metrics, but without a baseline, it's difficult to determine appropriate thresholds or if the current performance falls outside expected norms.

Forecasting uses historical data to predict future performance trends, which is useful for capacity planning but not for validating an immediate performance issue against past normal behavior.

Comparing current performance to a known historical baseline for that server and storage system allows for accurate identification of deviations and helps confirm if the observed slowdown is indeed a performance degradation.