White-box monitoring (also called glass-box monitoring) uses internal instrumentation to observe a system's behavior from the inside. It includes application metrics (request rates, error rates, queue depths), infrastructure metrics (CPU, memory, disk, network), application logs, and distributed traces.
White-box monitoring excels at root cause diagnosis. When black-box monitoring detects that a service is slow, white-box monitoring can reveal that it's because the database connection pool is exhausted, or a specific microservice is returning errors, or garbage collection pauses are causing latency spikes.
The limitation of white-box monitoring is that it can give a false sense of security — all internal metrics might look healthy while users are experiencing problems due to external factors (DNS, CDN, network routing). This is why best practice combines white-box monitoring for debugging and capacity planning with black-box monitoring (like Hyperping) for user-facing availability verification.