scadaproj/ZB.MOM.WW.Telemetry/README.md

# ZB.MOM.WW.Telemetry

Observability libraries for the **ZB.MOM.WW SCADA family** (OtOpcUa, MxAccessGateway, ScadaBridge). These are **libraries, not a service** — each package is linked directly into the consuming application at build time. There is no central telemetry process; all instrumentation runs in-process alongside the application.

The library normalizes the three-project observability surface: a shared OpenTelemetry Resource identity, standard instrumentation wiring, Prometheus and OTLP export, and a Serilog bootstrap with enrichers and trace↔log correlation — so metrics, traces, and log lines from the same node carry identical dimensions and can join up in any backend.

---

## Packages

| Package | Description | Key Dependencies |
|---|---|---|
| `ZB.MOM.WW.Telemetry` | `AddZbTelemetry` extension, `ZbTelemetryOptions`, shared OTel Resource builder (`ZbResource`), standard instrumentation (ASP.NET Core, HttpClient, gRPC client, runtime, process), Prometheus always-on exporter + OTLP opt-in overlay, `app.MapZbMetrics()` endpoint extension. | `Microsoft.AspNetCore.App` (framework ref), `OpenTelemetry.*` stack |
| `ZB.MOM.WW.Telemetry.Serilog` | `AddZbSerilog` extension, shared enrichers (`SiteId`/`NodeRole`/`NodeHostname`), `TraceContextEnricher` (writes `trace_id`/`span_id` from `Activity.Current` into every log event), `ILogRedactor` seam (per-project sensitive-field redaction), `RedactionEnricher`. | `ZB.MOM.WW.Telemetry`, `Serilog.*` stack |

---

## The unifying hinge

The single `ZbTelemetryOptions` object drives both packages. Its identity triple —
`ServiceName` → OTel Resource `service.name`, `SiteId` → `site.id`, `NodeRole` → `node.role` —
is applied once and flows automatically to **both** the OpenTelemetry Resource (so every metric
and span carries it) **and** the Serilog enrichers (so every log event carries it). A metric,
a span, and a log line emitted by the same node share identical `service.name`, `site.id`, and
`node.role` dimensions, enabling cross-signal correlation in any backend (Grafana, Jaeger, Seq,
Loki, etc.) without per-project bookkeeping.

---

## Consumer matrix

| Consumer | `ZB.MOM.WW.Telemetry` (core) | `ZB.MOM.WW.Telemetry.Serilog` |
|---|:---:|:---:|
| **OtOpcUa** | yes | yes |
| **MxAccessGateway** | yes | yes (logging adopted — MEL → Serilog migration done) |
| **ScadaBridge** | yes | yes |

All three apps consume both packages after adoption. MxAccessGateway's MEL→Serilog migration
is the one in-pass adoption completed on its own branch; OtOpcUa and ScadaBridge adoption is
follow-on (tracked in `components/observability/GAPS.md`).

---

## OTel signals

`AddZbTelemetry` wires all three OpenTelemetry signals in a single call:

| Signal | What is wired |
|---|---|
| **Metrics** | App Meters (via `options.Meters[]`) + standard: ASP.NET Core, HttpClient, .NET runtime, process. Exported via Prometheus (always on) with OTLP as an additive overlay. |
| **Traces** | App ActivitySources (via `options.ActivitySources[]`) + standard: ASP.NET Core, HttpClient, gRPC client. Exported via OTLP when `Exporter = ZbExporter.Otlp`. |
| **Logs** | Wired by `AddZbSerilog` (companion call). Serilog is used as the log sink; logs are bridged to OpenTelemetry via `Serilog.Sinks.OpenTelemetry` when configured. |

Trace↔log correlation is automatic: `TraceContextEnricher` reads `Activity.Current` for each
log event and attaches `trace_id` and `span_id`, so log events produced inside a traced request
carry the same span identity as the trace backend.

---

## Exporter options

Prometheus is **always wired** for metrics regardless of the `Exporter` setting. OTLP is an
additive overlay — set `Exporter = ZbExporter.Otlp` and `OtlpEndpoint` to push to a collector
in addition to the scrape endpoint.

```csharp
// Prometheus only (default — scrape /metrics)
builder.AddZbTelemetry(o =>
{
    o.ServiceName = "mxgateway";
    o.SiteId      = config["Site:Id"];
    o.NodeRole    = "standalone";
    o.Meters      = ["ZB.MOM.WW.MxGateway"];
});

// OTLP overlay (metrics + traces pushed to collector; /metrics still active)
builder.AddZbTelemetry(o =>
{
    o.ServiceName  = "mxgateway";
    o.SiteId       = config["Site:Id"];
    o.NodeRole     = "standalone";
    o.Meters       = ["ZB.MOM.WW.MxGateway"];
    o.Exporter     = ZbExporter.Otlp;
    o.OtlpEndpoint = "http://collector:4317";
});

// Mount the Prometheus scrape endpoint (call after app.UseRouting())
app.MapZbMetrics();   // → /metrics
```

```csharp
// Serilog bootstrap (same options object drives enrichers)
builder.AddZbSerilog(o =>
{
    o.ServiceName = "mxgateway";
    o.SiteId      = config["Site:Id"];
    o.NodeRole    = "standalone";
});
```

---

## Building and testing

```bash
# from ZB.MOM.WW.Telemetry/
dotnet build ZB.MOM.WW.Telemetry.slnx
dotnet test  ZB.MOM.WW.Telemetry.slnx
```

All test assemblies run with no external dependencies (no running OTel collector, no Serilog
backend):

| Assembly | Tests |
|---|---|
| `ZB.MOM.WW.Telemetry.Tests` | 7 |
| `ZB.MOM.WW.Telemetry.Serilog.Tests` | 12 |
| **Total** | **19** |

---

## Packing

```bash
dotnet pack ZB.MOM.WW.Telemetry.slnx -c Release -o ./artifacts
```

Produces two `.nupkg` files in `artifacts/`:

```
ZB.MOM.WW.Telemetry.0.1.0.nupkg
ZB.MOM.WW.Telemetry.Serilog.0.1.0.nupkg
```

`GeneratePackageOnBuild` is off — pack explicitly as above. Both packages are versioned
lockstep from `Directory.Build.props`.

---

## Status

**Built at 0.1.0. MxAccessGateway logging adopted (MEL → Serilog migration, on its own branch).
Broader OtOpcUa and ScadaBridge telemetry adoption deferred.** Adoption is tracked in the
component backlog:

- `~/Desktop/scadaproj/components/observability/GAPS.md`

Design documentation lives alongside that backlog:

- `~/Desktop/scadaproj/components/observability/spec/SPEC.md` — normalized observability target
- `~/Desktop/scadaproj/components/observability/spec/METRIC-CONVENTIONS.md` — metric naming reference
- `~/Desktop/scadaproj/components/observability/shared-contract/ZB.MOM.WW.Telemetry.md` — proposed API
- `~/Desktop/scadaproj/components/observability/current-state/` — per-project current state (code-verified)