Monitoring layers and KPIs

Source: Monitoring & Alerting, sys_kb_id=972394c02b80835ce9affd3fc891bf04.

---

The eight layers

#	Layer
1	Infrastructure
2	Network
3	SQL DB & Storage
4	Application
5	User Experience
6	Business
7	Security & Compliance
8	Disaster Recovery

[TBD: the standard does not enumerate the specific signals, metrics or thresholds required within each layer.]

What every layer requires

PagerDuty is the primary alerting path. Each of the eight layers requires:

1. Service(s) defined in PagerDuty
2. An escalation policy
3. Schedules

Important

All three, for every layer. A layer is not covered because an alert exists somewhere; it is covered when the alert reaches a defined PagerDuty service with an escalation policy and an on-call schedule behind it.

PagerDuty itself is approved but requires approval before use (Approved Software standard).

Universal KPIs

KPI	Meaning
MTTD	Mean time to detect
MTTA	Mean time to acknowledge
MTTR	Mean time to resolve

These are universal — they apply to every layer, not only to production incidents.

DORA metrics

Metric
Deployment frequency
Lead time for changes
MTTR
Change failure rate

Deployment frequency, lead time and change failure rate are produced by the CI/CD pipeline, which ties this standard to the CI/CD Pipeline Standards. MTTR is shared between the universal KPIs and DORA.

Warning

[TBD: no target values, thresholds or reporting frequency are stated for any KPI or DORA metric. Do not quote industry benchmarks as if they were Patterson targets.]

Review cadence

Annually. This is the only one of the six standards that states its own review cadence.

---

Source of truth: plugins/patterson-engineering/skills/monitoring-alerting-standards/references/layers-and-kpis.md in the patterson-corp repository.