This guide covers common issues encountered when deploying Flipt v2 to Kubernetes using the official Helm chart, along with their solutions.
Helm Schema Validation Errors
When installing or upgrading the Flipt Helm chart, you may encounter schema validation errors like:
helm install flipt flipt/flipt-v2 -f values.yaml
Error: INSTALLATION FAILED: values don't meet the specifications of the schema(s) in the following chart(s):
flipt-v2:
- autoscaling: Additional property targetMemoryUtilizationPercentage is not allowed
- resources: Additional property limits is not allowed
- resources: Additional property requests is not allowed
This happens when the chart’s JSON schema doesn’t cover all valid configuration options. To work around this, skip schema validation (requires Helm 3.13+):
helm install flipt flipt/flipt-v2 -f values.yaml --skip-schema-validation
Or for upgrades:
helm upgrade flipt flipt/flipt-v2 -f values.yaml --skip-schema-validation
--skip-schema-validation bypasses chart schema checks entirely. Use it as a
temporary workaround while tracking upstream schema fixes in
flipt-io/helm-charts — typos in
your values.yaml will no longer be caught automatically when this flag is
set.
Machine ID Not Found (License Validation)
When running Flipt Pro on containerd-based Kubernetes clusters (such as GKE or modern EKS), you may see:
license is invalid; additional features are disabled.
{"error": "machineid: machineid: no machine-id found"}
Why This Happens
Flipt uses a machine fingerprint for license validation. The fingerprinting library reads from:
/var/lib/dbus/machine-id
/etc/machine-id
- Docker-specific paths in
/proc/self/cgroup and /proc/self/mountinfo
In containerd-based environments (GKE, EKS with containerd, etc.), none of these files exist inside the container because:
- The Flipt container image is Alpine-based and doesn’t include dbus
/etc/machine-id is not present in minimal container images
- The Docker-specific fallback paths don’t match containerd’s format
Solution: Mount the Host Machine ID
Mount the Kubernetes node’s /etc/machine-id into the Flipt container by adding these values to your values.yaml:
extraVolumeMounts:
- name: machine-id
mountPath: /etc/machine-id
readOnly: true
extraVolumes:
- name: machine-id
hostPath:
path: /etc/machine-id
This mounts the host node’s machine ID into the container, allowing the license system to identify the machine.
Affected Environments
- Google Kubernetes Engine (GKE)
- Amazon EKS with containerd runtime
- Any Kubernetes cluster using containerd instead of Docker
- Rootless container environments
- Amazon EKS with AWS Fargate node groups
Persistent Volume Configuration
Flipt needs a writable directory for local state. If you see read-only filesystem errors or Flipt fails to start with write permission errors, you likely need to configure persistence.
Enable Persistence
Add the following to your values.yaml:
persistence:
enabled: true
storageClass: "standard" # Use your cluster's storage class
size: 10Gi
Run kubectl get storageclass to see available storage classes in your
cluster. The default varies by provider and cluster configuration.
Match the Storage Backend Path
The path in your Flipt storage backend configuration refers to the path inside the container, not on the host. It must match where the PersistentVolumeClaim is mounted.
The Helm chart mounts the PVC at /var/opt/flipt by default. Your storage backend configuration should use this path:
flipt:
config:
storage:
my-storage:
backend:
type: local
path: /var/opt/flipt
Do not use a relative path like "." for the backend path. This writes to the
container’s working directory, which may be read-only. Always use the absolute
path where the PVC is mounted.
Secrets Management with GitOps
When deploying Flipt via ArgoCD, FluxCD, or other GitOps tools, you should not commit sensitive values (license keys, Git access tokens) to your Git repository inside values.yaml.
Using Kubernetes Secrets with Environment Variable Substitution
Flipt supports environment variable substitution in configuration values using the ${env:VAR_NAME} syntax. Combined with the Helm chart’s envFrom support, this lets you keep secrets out of your values file.
Step 1: Create a Kubernetes Secret containing your sensitive values:
kubectl create secret generic flipt-secrets \
--from-literal=FLIPT_LICENSE_KEY=your-license-key \
--from-literal=FLIPT_GIT_ACCESS_TOKEN=your-access-token
Or use External Secrets Operator to sync from your cloud provider’s secret manager (GCP Secret Manager, AWS Secrets Manager, Azure Key Vault, HashiCorp Vault).
Step 2: Reference the Secret in your values.yaml:
flipt:
envFrom:
- secretRef:
name: flipt-secrets
Step 3: Use environment variable references in your Flipt configuration:
flipt:
config:
license:
key: ${env:FLIPT_LICENSE_KEY}
credentials:
my-git:
type: access_token
access_token: ${env:FLIPT_GIT_ACCESS_TOKEN}
storage:
my-storage:
remote: "https://github.com/your-org/flipt-config.git"
branch: main
credentials: "my-git"
This approach keeps your values.yaml free of secrets and safe to commit to Git.
This pattern works with any tool that creates Kubernetes Secrets:
- External Secrets Operator — syncs secrets from GCP Secret Manager, AWS Secrets Manager, Azure Key Vault, HashiCorp Vault
- Sealed Secrets — encrypt secrets client-side, commit encrypted
SealedSecret resources to Git
- ArgoCD Vault Plugin — inject secrets from Vault at deploy time
kubectl create secret — create secrets manually for simpler setups
Using Individual Environment Variables
If you prefer to inject secrets as individual environment variables rather than from a Secret reference, use extraEnvVars:
flipt:
extraEnvVars:
- name: FLIPT_LICENSE_KEY
valueFrom:
secretKeyRef:
name: flipt-license
key: license-key
- name: FLIPT_GIT_ACCESS_TOKEN
valueFrom:
secretKeyRef:
name: flipt-git-creds
key: access-token