No dedicated repository. This case documents the observability-mesh pattern; production governance that consumes those signals lives in the AI Infra Control Plane.
Case 10 · Pattern
OpenTelemetry Observability Mesh
An observability mesh connects traces, metrics, logs, runbooks, and ownership so production behavior can be understood from user request to Kubernetes workload.
00
Evidence home
Pattern case — proof lives next door.
01
How it works
One request — a river of spans.
OpenTelemetry Observability Mesh by Andrey Lesnikov — connect traces, metrics, logs, and runbooks so incidents tell one story.
-
01
Emit the path
Services paint spans in order — gateway to dependency — so the waterfall is evidence, not decoration.
-
02
Collect without chaos
Collectors normalize and sample so cardinality stays a budget, not an accident.
-
03
Correlate the signals
Logs and metrics hitch to trace context — one investigation flow instead of three dashboards.
-
04
Land on an owner
SLO links and runbooks turn a red chart into a named human and a next step.
02
Evidence
From problem to proof.
How the system is framed, operated, and proven — not a slide deck summary.
01
Incidents still require stitching separate dashboards: traces, metrics, logs, and runbooks do not tell one request story.
02
You mainly need SLO burn-rate language or AI token cost — use SLO-Driven Monitoring or Cost / Token Observability.
03
SLO-Driven Monitoring turns correlated signals into reliability decisions. AI Infra Control Plane consumes the same observability fabric for governance.
Constraints
Cardinality, sampling, cost, multi-service correlation, dashboard sprawl, and alert fatigue.
03
Related
Nearby systems.
SLO-Driven Monitoring
Dashboards can look healthy while users experience latency, errors, or degraded workflows
Alerts from SLO burn, not chart noise
Cost and Token Observability
AI and cloud costs can grow quietly when usage is disconnected from teams, services, and deployment changes
Token spend visible before finance week
AI Infra Control Plane
Private AI platforms need shared governance across identity, policy, audit, cost, and SLO signals — not scattered service configs
Private AI with explicit platform contracts
Available for meaningful infrastructure conversations