HPA was pinning the deployment at maxReplicas (4) even though aggregate
CPU usage was low. The service is event-driven: pods sit at ~0-1m idle
but spike to 100-300m per supergraph query. With requests.cpu=20m those
bursts read as 500-1500% utilization, so the HPA's 60% target was
constantly exceeded on whichever pod handled the request.
Raise the request to 100m so bursts read as 100-300% instead. Combined
with the HPA's downscale stabilization window this lets replicas settle
back to minReplicas (2) between bursts.
The receiver suffix was stale — the `grafana-k8s-monitoring` module deploys a single Service named `k8s-monitoring-alloy` in the `monitoring` namespace, not `k8s-monitoring-alloy-receiver`. Pods were failing to ship traces with:
```
dial tcp: lookup k8s-monitoring-alloy-receiver.monitoring.svc.cluster.local on 100.64.0.10:53: no such host
```
Drop the `-receiver` suffix so the OTLP exporter resolves the correct service.
Reviewed-on: #839
- Update git remote to git.unbound.se
- Add Gitea workflows: ci.yaml, pre-commit.yaml, release.yaml, goreleaser.yaml
- Delete .gitlab-ci.yml
- Update Go module path to gitea.unbound.se/unboundsoftware/schemas
- Update all imports to new module path
- Update Docker registry to oci.unbound.se
- Update .goreleaser.yml for Gitea releases with internal cluster URL
- Remove GitLab CI linter from pre-commit config
- Use shared release workflow with tag_only for versioning
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Introduce health checking functionality with liveness and readiness
endpoints to monitor the application's status. Implement a health
checker that verifies database connectivity and provides a simple
liveness check. Update service routing to use the new health
checker functionality. Add corresponding unit tests for validation.
Add the OTEL_EXPORTER_OTLP_ENDPOINT environment variable to
the deployment configuration. This change enables the application
to send telemetry data to the specified OpenTelemetry collector
endpoint for better monitoring and observability.
Remove Sentry dependencies and configuration. Introduce monitoring
setup for OpenTelemetry. Update logging to include log format
options, and replace Sentry error handling middleware with
monitoring handlers for GraphQL playground. Adjust environment
variable handling to enhance configuration clarity and flexibility.
Removes the CPU and memory limits for the schemas container in the
Kubernetes deployment configuration. This change is made to allow
for greater flexibility in resource allocation based on runtime
demands, ensuring better performance in varying load conditions.
Add a new extract for RabbitMQ in the secrets and include
AMQP_URL in local secrets. Update the merge policy and
engine version in the deployment template. Remove unused
RabbitMQ secret reference from the deployment. These changes
improve configuration management and ensure consistency
between environments.
Update Kubernetes YAML files to replace the app label with the
standard app.kubernetes.io/name label for better consistency
and adherence to best practices in Kubernetes resource labeling.
Reduces the CPU request from "100m" to "10m" in the deploy.yaml file to
optimize resource usage and improve deployment efficiency in the
Kubernetes environment. This change helps to allocate resources more
effectively while maintaining application performance.
Adds a package rule to disable updates for the
'unboundsoftware/schemas' package in Kubernetes.
Modifies the image reference in the deployment
configuration to remove the specific SHA256 hash
and use a dynamic commit tag instead. This
improves maintainability and aligns with the
latest deployment practices.
argoyle
Renovate