- Added detailed task completion records for KMS interface implementation and CLI support for file-based keys. - Documented security enhancements including Argon2id password hashing, audit event contracts, and rate limiting configurations. - Included scoped service support and integration updates for the Plugin platform, ensuring proper DI handling and testing coverage.
1.8 KiB
Telemetry Storage Stack
Configuration snippets for the default StellaOps observability backends used in staging and production environments. The stack comprises:
- Prometheus for metrics (scraping the collector's Prometheus exporter)
- Tempo for traces (OTLP ingest via mTLS)
- Loki for logs (HTTP ingest with tenant isolation)
Files
| Path | Description |
|---|---|
prometheus.yaml |
Scrape configuration for the collector (mTLS + bearer token placeholder). |
tempo.yaml |
Tempo configuration with multitenancy enabled and local storage paths. |
loki.yaml |
Loki configuration enabling per-tenant overrides and boltdb-shipper storage. |
tenants/tempo-overrides.yaml |
Example tenant overrides for Tempo (retention, limits). |
tenants/loki-overrides.yaml |
Example tenant overrides for Loki (rate limits, retention). |
auth/ |
Placeholder directory for Prometheus bearer token files (e.g., token). |
These configurations are referenced by the Docker Compose overlay
(deploy/compose/docker-compose.telemetry-storage.yaml) and the staging rollout documented in
docs/modules/telemetry/operations/storage.md. Adjust paths, credentials, and overrides before running in
connected environments. Place the Prometheus bearer token in auth/token when using the
Compose overlay (the directory contains a .gitkeep placeholder and is gitignored by default).
Run python ops/devops/telemetry/validate_storage_stack.py after editing any of these files to
ensure TLS, multitenancy, and override references remain intact.
Security
- Both Tempo and Loki require mutual TLS.
- Prometheus uses mTLS plus a bearer token that should be minted by Authority.
- Update the overrides files to enforce per-tenant retention/ingestion limits.
For comprehensive deployment steps see docs/modules/telemetry/operations/storage.md.