L-għan huwa li titqabbaḍ l-issues qabel il-paġna — biex issib degradazzjoni filwaqt li hemm għadha slack biex tiassorbih. Dan ifisser li tħares leading indicators, tiddefinixxi SLOs b'error budgets, u tattestja attivament is-sistema billi tistenna li tfalli.
An SLO idawwal ir-reliability f'numru (per eż. 99.9% tar-requests jirnexxu). L-ibqija 0.1% hija l-error budget tiegħek. Tracking tal-burn rate ilaqqgħok tialerta meta tkun qed tinfaq il-budget malajr wisq — ħafna qabel ma tkun attwalment break l-SLO u l-utenti jinnotu.
SLO 99.9% → 0.1% error budget/month (~43 min of downtime)
burn rate rising fast → you'll exhaust it in 2 days → alert NOW, while it's fixable
SYNTHETIC MONITORING scripted checks hit critical paths on a schedule
(login, checkout) → fails even at 3am with zero real traffic
HEALTH CHECKS /healthz endpoints + dependency checks → load balancer
pulls bad instances before users hit them
RUM (real-user mon.) measure latency/errors from actual browsers/devices →
catches issues only some users/regions see
Synthetic monitoring hija qawwija għaliex ma tistennax user — ikkuntinwament tesercizza s-sistema, għalhekk checkout miksur jinstab f'3am, mhux meta l-midnight rush jikomplew.
L-ewwel sinjali huma f'resources, għadha mhux f'user-facing errors. Alerta fuq it-trend, mhux biss line statiku.
LEADING INDICATORS saturation (CPU/mem climbing), queue depth growing,
connection-pool nearing limit, latency CREEPING up
ANOMALY DETECTION flag deviation from the normal baseline / seasonality
TREND ALERTS "disk will fill in 4h at this rate" → act before it's full
P99 li qed jogħla bil-mod jew queue li qed tikber hija warning shot: billi taġixxi fuq il-creep, tipprevieni l-outage li l-creep kien qed jimxi lejnha.
Reactive monitoring ifisser li l-utenti huma s-sistema ta' alerting tiegħek — meta jikomplew, l-incident huwa diġà live u l-error budget tiegħek huwa miċħud. Proactive detection (SLO burn rate, synthetics, health checks, RUM, leading indicators, trend/anomaly alerts) tiegħti lead time: inti tissoṭṭa queue li qed tissatura jew latency li qed tikber qabel ma ssir 2am page u customer milħun. Dak il-lead time huwa d-differenza bejn fix silenzjuż u outage.