feat(telemetry): add Phase 5 documentation, deployment configs, and integration tests

Add the observability stack deployment infrastructure and integration
test framework for verifying end-to-end trace export.

- Add Grafana dashboards: RPC performance, transaction overview,
  consensus health (pre-provisioned via dashboards.yaml)
- Add Prometheus config for spanmetrics collection from OTel Collector
- Update OTel Collector config with spanmetrics connector and
  prometheus exporter for RED metrics
- Add docker-compose services: prometheus, dashboard provisioning
- Add integration-test.sh with Tempo API-based span verification
  (replaces previous Jaeger-based approach)
- Add TESTING.md with step-by-step deployment and verification guide
- Add telemetry-runbook.md for production operations reference
- Add xrpld-telemetry.cfg sample configuration
- Add toDisplayString() for ConsensusMode (human-readable span values)
- Update Phase 2/3 task lists with known issues sections
- Add Phase 5 integration test task list
- Add TraceContext protobuf fields for future relay propagation
- Wire telemetry lifecycle (setServiceInstanceId/start/stop) in
  Application.cpp

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

OpenTelemetryPlan/Phase2_taskList.md

@@ -204,3 +204,36 @@ This surfaces all RPCs served during a blocked period — critical for post-inci
 **Delivered in this branch**: Tasks 2.4, 2.7, 2.8, 2.9.
 **Deferred with rationale**: Tasks 2.1 (→Phase 3), 2.5 (low priority).
 **Superseded**: Task 2.2 (Phase 1c SpanGuard factory covers this).
+
+---
+
+## Known Issues / Future Work
+
+### Thread safety of TelemetryImpl::stop() vs startSpan()
+
+`TelemetryImpl::stop()` resets `sdkProvider_` (a `std::shared_ptr`) without
+synchronization. `getTracer()` reads the same member from RPC handler threads.
+This is a data race if any thread calls `startSpan()` concurrently with `stop()`.
+
+**Current mitigation**: `Application::stop()` shuts down `serverHandler_`,
+`overlay_`, and `jobQueue_` before calling `telemetry_->stop()`, so no callers
+remain. See comments in `Telemetry.cpp:stop()` and `Application.cpp`.
+
+**TODO**: Add an `std::atomic<bool> stopped_` flag checked in `getTracer()` to
+make this robust against future shutdown order changes.
+
+### Macro incompatibility: XRPL_TRACE_SPAN vs XRPL_TRACE_SET_ATTR
+
+`XRPL_TRACE_SPAN` and `XRPL_TRACE_SPAN_KIND` declare `_xrpl_guard_` as a bare
+`SpanGuard`, but `XRPL_TRACE_SET_ATTR` and `XRPL_TRACE_EXCEPTION` call
+`_xrpl_guard_.has_value()` which requires `std::optional<SpanGuard>`. Using
+`XRPL_TRACE_SPAN` followed by `XRPL_TRACE_SET_ATTR` in the same scope would
+fail to compile.
+
+**Current mitigation**: No call site currently uses `XRPL_TRACE_SPAN` — all
+production code uses the conditional macros (`XRPL_TRACE_RPC`, `XRPL_TRACE_TX`,
+etc.) which correctly wrap the guard in `std::optional`.
+
+**TODO**: Either make `XRPL_TRACE_SPAN`/`XRPL_TRACE_SPAN_KIND` also wrap in
+`std::optional`, or document that `XRPL_TRACE_SET_ATTR` is only compatible with
+the conditional macros.