Monitoring, Analytics, and Assurance
From polling to telemetry
Know the difference between SNMP polling (periodic, coarse) and streaming telemetry (continuous, high-resolution, push-based). Understand what each layer measures: infrastructure health (CPU, interface errors, PoE), flow data (NetFlow/IPFIX), wireless RF metrics (channel utilization, retries, SNR, client experience), and application experience (synthetic tests, latency, jitter, MOS). AIOps-style analytics establish baselines and surface anomalies rather than relying only on static thresholds.
Assurance and SLAs
Assurance means proving the network delivers the intended experience. Know how to define service-level objectives, measure them with the right probe placement, and distinguish a symptom from a root cause. Wireless assurance often scores client experience from onboarding through roaming to application use; wired assurance focuses on interface health, PoE, and path performance.
Traps
Common traps: monitoring a link's utilization but not its error/discard counters; placing probes where they cannot see the user-to-application path; assuming a green dashboard means the user experience is good; and alerting on every threshold instead of actionable conditions. Also watch for questions that confuse flow records (metadata) with packet capture (payload).
Worked example
Users report slow Wi-Fi in a conference room. Rather than raising AP power, check channel utilization, retry rate, client SNR, and the number of clients per AP during the meeting. If utilization is high but SNR is good, the issue is capacity/co-channel interference, not coverage. The fix is channel and AP density planning, not more transmit power, which would worsen interference.