☸️ Kubernetes & Helm Deployment¶

Developer-friendly, production-ready Helm chart for deploying Spector on Kubernetes.

Spector packages an official Helm chart published to the GitHub Container Registry as an OCI artifact (oci://ghcr.io/spectrayan/charts/spector).

Quick Start¶

1. Install from GHCR (OCI)¶

No external repository setup is needed. Install directly using Helm 3.8+:

helm install spector oci://ghcr.io/spectrayan/charts/spector \
  --namespace spector \
  --create-namespace

2. Verify Deployment¶

kubectl get pods -n spector
kubectl get pvc -n spector

Port-forward to access Synapse locally:

kubectl port-forward svc/spector 7070:7070 -n spector

Architecture & Workload Layout¶

The Helm chart provisions: - StatefulSet: Ensures persistent node identity and guarantees safe storage unmounting during rolling upgrades. - PersistentVolumeClaim: Dedicated block or network storage mounted at /data for WAL and off-heap memory segments. - Service: Internal ClusterIP on port :7070 with readiness and liveness HTTP probes. - ConfigMap: Application properties configuring embedders, storage paths, and logging.

Local Cluster Profiles (kind / k3d / Minikube)¶

For local development clusters with limited memory:

helm install spector oci://ghcr.io/spectrayan/charts/spector \
  --namespace spector \
  --create-namespace \
  --set persistence.size=5Gi \
  --set resources.requests.cpu=250m \
  --set resources.requests.memory=1Gi \
  --set resources.limits.memory=2Gi

Configuration Reference (`values.yaml`)¶

Key parameters in deploy/helm/spector/values.yaml:

Parameter	Default	Description
`image.repository`	`ghcr.io/spectrayan/spector`	Container image repository
`image.tag`	`0.1.0-alpha`	Container tag
`service.type`	`ClusterIP`	Kubernetes service type (`ClusterIP`, `NodePort`, `LoadBalancer`)
`service.port`	`7070`	Service port
`persistence.enabled`	`true`	Enable persistent storage via PVC
`persistence.size`	`10Gi`	Disk storage allocation
`persistence.storageClass`	`""`	Storage class name (empty string uses cluster default)
`resources.requests.memory`	`1Gi`	Minimum memory request
`resources.limits.memory`	`4Gi`	Maximum memory limit
`spector.embedderProvider`	`onnx`	Default embedder provider (`onnx`, `ollama`, `openai`)

Upgrading a Release¶

helm upgrade spector oci://ghcr.io/spectrayan/charts/spector \
  --namespace spector \
  --values my-values.yaml

Uninstalling¶

helm uninstall spector --namespace spector

(Note: PersistentVolumeClaims are retained by default to prevent accidental data loss.)