mirror of
https://github.com/XRPLF/rippled.git
synced 2026-03-13 16:22:31 +00:00
4db67bc191
Add comprehensive workload harness for end-to-end validation of the
Phases 1-9 telemetry stack:
Task 10.1 — Multi-node test harness:
- docker-compose.workload.yaml with full OTel stack (Collector, Jaeger,
Tempo, Prometheus, Loki, Grafana)
- generate-validator-keys.sh for automated key generation
- xrpld-validator.cfg.template for node configuration
Task 10.2 — RPC load generator:
- rpc_load_generator.py with WebSocket client, configurable rates,
realistic command distribution (40% health, 30% wallet, 15% explorer,
10% tx lookups, 5% DEX), W3C traceparent injection
Task 10.3 — Transaction submitter:
- tx_submitter.py with 10 transaction types (Payment, OfferCreate,
OfferCancel, TrustSet, NFTokenMint, NFTokenCreateOffer, EscrowCreate,
EscrowFinish, AMMCreate, AMMDeposit), auto-funded test accounts
Task 10.4 — Telemetry validation suite:
- validate_telemetry.py checking spans (Jaeger), metrics (Prometheus),
log-trace correlation (Loki), dashboards (Grafana)
- expected_spans.json (17 span types, 22 attributes, 3 hierarchies)
- expected_metrics.json (SpanMetrics, StatsD, Phase 9, dashboards)
Task 10.5 — Performance benchmark suite:
- benchmark.sh for baseline vs telemetry comparison
- collect_system_metrics.sh for CPU/memory/latency sampling
- Thresholds: <3% CPU, <5MB memory, <2ms RPC p99, <5% TPS, <1% consensus
Task 10.6 — CI integration:
- telemetry-validation.yml GitHub Actions workflow
- run-full-validation.sh orchestrator script
- Manual trigger + telemetry branch auto-trigger
Task 10.7 — Documentation:
- workload/README.md with quick start and tool reference
- Updated telemetry-runbook.md with validation and benchmark sections
- Updated 09-data-collection-reference.md with validation inventory
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>