Skip to content

Helm Values Reference

The Sidereal Helm chart is the primary deployment mechanism. All operational behavior is configurable through values.yaml, with a JSON schema (values.schema.json) enforcing validation constraints.

Top-level settings that cascade defaults across the deployment.

KeyTypeDefaultDescription
global.impactLevelhigh|moderate|lowhighFIPS 199 impact level. Cascades cadence, retention, and fail-closed defaults.
global.executionModedryRun|observe|enforcedryRunDefault execution mode for all probes.
global.fipsbooltrueRequire FIPS 140-2 validated cryptography (BoringCrypto).
global.requireAdmissionControllerbooltrueRequire a supported admission controller.
global.controlFrameworkslist["nist-800-53"]Compliance frameworks to load crosswalk files for.

Impact level defaults:

  • high: 6-hour probe cadence, 365-day retention, fail-closed on export failure
  • moderate: 12-hour cadence, 365-day retention, no fail-closed
  • low: 24-hour cadence, 180-day retention, no fail-closed

Selects backend integrations. Sidereal references abstract capabilities, not specific tools. Six pre-built profiles ship with the chart.

KeyOptionsDefault
profile.admissionControllerkyverno, opakyverno
profile.signatureVerifierkyverno, policy-controllerkyverno
profile.detectionBackendfalco, tetragon, nonefalco
profile.cniObservabilityhubble, calico, tcp-inferencetcp-inference

Controller Manager deployment settings.

KeyDefaultDescription
controller.image.repositoryghcr.io/primaris-tech/sidereal-controllerController image.
controller.image.tag"" (appVersion)Image tag.
controller.replicas1Replica count.
controller.resources.requests.cpu100mCPU request.
controller.resources.requests.memory128MiMemory request.
controller.resources.limits.cpu500mCPU limit.
controller.resources.limits.memory512MiMemory limit.

Probe runner configuration.

KeyDefaultDescription
probe.intervalSeconds21600Default interval between executions (overridden per-probe). Schema enforces 300-86400.
probe.goImage.repositoryghcr.io/primaris-tech/sidereal-probe-goGo probe runner image.
probe.detectionImage.repositoryghcr.io/primaris-tech/sidereal-probe-detectionRust detection probe image.
probe.bootstrapImage.repositoryghcr.io/primaris-tech/sidereal-probe-bootstrapBootstrap verification image.

Audit record retention and fail-closed behavior.

KeyDefaultDescription
audit.retentionDays365Minimum retention for probe results.
audit.failClosedOnExportFailuretrueHalt probe scheduling if SIEM export fails consecutively.

SIEM export target configuration. Supports multiple simultaneous targets.

export:
targets:
- backend: splunk|elasticsearch|s3
format: json|cef|leef|syslog|ocsf
endpoint: https://...
secretRef: credential-secret-name
bucket: bucket-name # S3 only
region: us-east-1 # S3 only

Detection backend endpoints, used when profile.detectionBackend is set.

KeyDefault
detection.falco.endpointfalco-grpc.falco:50051
detection.tetragon.endpointtetragon.kube-system:54321

CNI observability endpoints, used when profile.cniObservability is hubble or calico.

KeyDefault
cni.hubble.endpointhubble-relay.kube-system:4245
cni.calico.endpointhttps://calico-api.calico-system:5443
KeyDefaultDescription
tls.requiredtrueRequire TLS for all backend connections. Schema enforces this must be true.

Register ServiceAccounts for custom probe types.

customProbes:
serviceAccounts:
- name: my-custom-probe
namespace: sidereal-system

Caps concurrent probe Jobs in the sidereal-system namespace.

KeyDefaultDescription
resourceQuota.maxJobs10Maximum concurrent probe Jobs.
resourceQuota.maxCPU"4"Total CPU limit for probe Jobs.
resourceQuota.maxMemory8GiTotal memory limit for probe Jobs.