Which counters are useful when reading interface counters to diagnose issues?

• A. Jitter and packet size.
• B. Packet rate and queue depth.
• C. Total drops, errors, discards, alignment, and FEC corrected/uncorrected
• D. CPU load.

Answer: (C)

When diagnosing interface problems, you want counters that reveal loss and data integrity issues directly on the link. The set that matters most is total drops, errors, discards, alignment, and FEC corrected/uncorrected. These metrics tell you if packets aren’t getting through, if there are framing or physical-layer problems, and whether any error-correction is stepping in or failing. Drops show actual loss, errors indicate bit or frame problems, discards point to packets being thrown away by the interface, alignment flags framing issues, and FEC counters show how often errors are being corrected (or not). Together, they give a clear picture of a faulty or congested link and guide you toward problems like bad cabling, misconfiguration, or degraded physical hardware.

Other metrics don’t map as directly to the health of the link. Jitter and packet size describe timing and traffic characteristics rather than the presence of errors on the interface. Packet rate and queue depth can suggest congestion, but they don’t inherently indicate errors on the interface itself. CPU load is a system-level resource metric, not an interface counter, so it doesn’t reveal the state of the network interface’s transmission quality.