Kepriye carane ndeteksi masalah sadurunge panganggo ngeluh?

Question

Accepted Answer

Tujuane yaiku nangkep masalah **sadurunge nang kaca** — nemokake degradasi nalika isih ana ruang kanggo nyerap. Kuwi tenan mbutuhake nonton **indikator ngarep**, nemtokake **SLO karo error budget**, lan aktif ndeteksi sistem tinimbang ngenteni sistem rusak.

## SLO lan error budget

SLO **SLO** ngowahi reliabilitas dadi nomer (contone 99.9% saka request sukses). Sisa 0.1% iku **error budget** mu. Ntrack **burn rate** ngidini kowe alert nalika kowe mbuwang budget ketulungan cepet — suwe-suwe sadurunge bener-bener nglanggar SLO lan panganggo ngrasa.

```text
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
```

## Ndeteksi aktif, ora mung metrik pasif

```text
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 kuwi kuat amerga ora ngenteni panganggo — terus-terusan ngetung sistem, dadi checkout rusak ketemu jam 3 subuh, ora nalika sumurup pagi ngeluh.

## Indikator ngarep lan trend

Tanda paling awal ana ing resource, durung ana ing kesalahan ngarep panganggo. Alert ing **trend**, ora mung garis statis.

```text
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 naik alon-alon utawa antrian gedhe yaiku peringatan: kanthi tumindak ing nyampur, kowe nyegah outage sing nyampur mau arep menyang.

## Apa iku penting

Monitoring reaktif tegese panganggo dadi sistem alerting mu — nalika wis ngeluh, incident wis urip lan error budget mu wis ilang. Deteksi proaktif (SLO burn rate, synthetics, health check, RUM, indikator ngarep, trend/anomaly alert) tuku wektu ngarep: kowe ndandani antrian jenuh utawa latency nyampur sadurunge dadi 2am page lan panganggo nesu. Wektu ngarep kuwi beda antarane ndandani senyep lan outage.