All notable changes to this project will be documented in this file.
The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.
- OpenAPI schema is now sourced from the versioned
hyperfleet-api-specGo module instead of being downloaded fromhyperfleet-apimain branch
- BREAKING CHANGE:
messaging_systemconfig field andMESSAGING_SYSTEMenv var removed,messaging.systemOTel span attribute is now derived frompublisher.BrokerType(). Removemessaging_systemfrom configs before upgrading.
0.1.1 - 2026-03-10
- Standard metrics labels to Sentinel Prometheus metrics for consistent monitoring across HyperFleet components
- ServiceMonitor resource for Prometheus Operator environments
- PodDisruptionBudget to protect Sentinel availability during voluntary disruptions
- Helm chart linting and template validation to CI via Makefile targets
- Support for nested field paths in
message_dataconfiguration for richer event content - Functional health and readiness probes beyond basic liveness checks
- Updated
hyperfleet-brokerto v1.1.0 and integratedMetricsRecorderfor broker-level observability - Standardized Helm value structure for consistency across HyperFleet charts
- Moved Sentinel Helm chart to
charts/directory following repository conventions - GCP-specific monitoring resources are now disabled by default
- Standardized Dockerfile and Makefile for unified image build process
- Standardized version injection to avoid collisions with
go-toolsetenvironment variables
- RabbitMQ connection URL now included in broker ConfigMap for proper broker discovery
- CA certificates copied from builder stage to
ubi9-microruntime, resolving TLS verification failures - Clarified Helm deployment instructions for GKE environments using Quay images
0.1.0 - 2026-02-19
- Initial release of HyperFleet Sentinel Service
- Kubernetes resource polling for clusters and nodepools
- CloudEvents publishing with broker abstraction (GCP Pub/Sub, RabbitMQ, Stub)
- Horizontal sharding via resource selector labels
- Configurable polling intervals and max age intervals (not ready vs ready resources)
- CEL-based message data templating for custom CloudEvents payloads
- Prometheus metrics for observability
- Grafana dashboard for monitoring
- PodMonitoring support for GKE with Google Cloud Managed Prometheus
- Helm chart for deployment
- Integration tests with testcontainers (RabbitMQ and GCP Pub/Sub)
- OpenAPI client generation from hyperfleet-api specification
- Configuration validation at startup
- HyperFleet API client with retry logic
- Comprehensive test coverage and linting