Observability¶
Observability is the ability to understand a system's internal state by examining its external outputs. Unlike traditional monitoring that tells you when something is wrong, observability helps you understand why - even for problems you've never seen before.
IAPM brings observability into 3D space. Instead of switching between dashboards, walk through your service topology, teleport to services, and ask Tessa to explain what's happening.
The Three Pillars¶
graph TD
subgraph "The Three Pillars"
L[Logs]
M[Metrics]
T[Traces]
end
L --> O[IAPM - Correlated Observability]
M --> O
T --> O Traces¶
Records of requests as they flow through distributed systems. In IAPM, traces are visualized as 3D structures in the Diagnostics Room - you can walk through a request's journey across services.
Best for: Understanding request flow, identifying latency bottlenecks, debugging distributed failures.
Metrics¶
Numeric measurements collected at regular intervals - counters, gauges, and histograms. IAPM renders metrics as real-time graphs on the Grid surface.
Best for: Tracking trends, setting alert thresholds, capacity planning.
Logs¶
Timestamped records of discrete events. IAPM correlates logs with their parent traces, so you can jump from a failed span directly to the error log.
Best for: Debugging specific errors, audit trails, understanding application behavior at a granular level.
How the Pillars Work Together in IAPM¶
The real power is correlation - connecting signals across pillars:
| You Notice | Start With | Then Use | IAPM Experience |
|---|---|---|---|
| "Response times are slow" | Metrics (latency graph) | Traces (find slow spans) | Ask Tessa: "Why is the order service slow?" |
| "Error rate is spiking" | Metrics (error rate alert) | Logs (error details) | Tessa correlates alerts with root cause |
| "Request failed" | Traces (failed span) | Logs (exception stack trace) | Click the span, see the log inline |
OpenTelemetry¶
IAPM uses OpenTelemetry as its sole instrumentation framework. One SDK captures all three pillars - no separate agents for logs vs traces vs metrics.
Next Steps¶
- Learn how to instrument your application with OpenTelemetry
- Understand how Collection gathers telemetry data
- See how Correlation connects signals across pillars