Dash0 acquires Polar Signals

Last updated: October 8, 2026

Dash0 Synthetics Worker

Run Dash0 synthetic checks against private locations from inside your own Kubernetes cluster.

The synthetics-worker runs Dash0 synthetic checks from inside your own Kubernetes cluster, so that checks can reach internal services that Dash0's public locations cannot. It dials outbound only (no inbound ports are opened), polls Dash0 for work, and executes synthetic checks against a Dash0-issued private location.

A cluster can shard checks across multiple Dash0 private locations: each private location is configured as one instance, and the operator runs one Kubernetes Deployment per instance.

Quickstart

Enabling the synthetics-worker takes two steps: turning the feature flag on in the Helm chart, and configuring at least one instance on the Dash0OperatorConfiguration resource.

  1. Install or upgrade the Helm chart with the feature enabled and the Dash0 synthetics backend address set:

    shell
    1234
    helm upgrade --install dash0-operator dash0-operator/dash0-operator \
      --namespace dash0-system \
      --set operator.syntheticsWorker.enabled=true \
      --set operator.syntheticsWorker.serverAddress=ingress.<cell>.dash0.com:443

    The exact serverAddress for your organization is shown in the Dash0 UI when you create a private location.

  2. Configure at least one instance on your Dash0OperatorConfiguration resource. The private location ID and the authorization token are per-cluster settings that live on this resource (not on the Helm chart), so that they can be changed at runtime without a helm upgrade:

    yaml
    1234567891011
    apiVersion: operator.dash0.com/v1alpha1
    kind: Dash0OperatorConfiguration
    metadata:
      name: dash0-operator-configuration-resource
    spec:
      # ... your existing exports, etc. ...
      syntheticsWorker:
        instances:
          - locationId: my-cluster-location
            authorization:
              token: auth_...

    locationId identifies which Dash0 private location this instance serves checks for, and is also used to derive the names of the Kubernetes resources the operator creates for it. It must be unique within instances.

Please refer to operator_configuration_types.go for the full SyntheticsWorker/SyntheticsWorkerInstance schema, including per-instance replicas, resources, nodeAffinity and tolerations.

How it works

For each configured instance, the operator creates one Deployment and one ServiceAccount in the operator's own namespace. The authorization token is passed to the pod as an environment variable, either from the literal authorization.token value or, via authorization.secretRef, from a Kubernetes secret — the operator never reads that secret itself; Kubernetes substitutes it into the pod's environment directly.

Every instance's NodeAffinity defaults to the same node selection used by the operator's other workloads (excluding nodes labeled dash0.com/enable=false, and requiring kubernetes.io/os=linux) unless you set nodeAffinity explicitly on that instance.

Opting Out

Two independent opt-outs exist:

  • Per-cluster, at runtime: set spec.syntheticsWorker.enabled: false on the Dash0OperatorConfiguration resource. The operator removes every synthetics-worker Deployment and ServiceAccount it manages. Re-enabling (or removing the field, since it defaults to true) redeploys them.
  • At the Helm level: set operator.syntheticsWorker.enabled=false and run helm upgrade. This disables the feature chart-wide; note that the Helm chart does not remove already-deployed synthetics-worker resources on its own when you flip this flag — uninstall and reinstall the operator if you need them removed immediately.

Status

The operator reports the aggregate and per-instance state on Dash0OperatorConfiguration.status.syntheticsWorker: deployed/reason/message for whether every instance was successfully created or updated, and ready/readyReason for whether every instance has all of its desired replicas ready. status.syntheticsWorker.instances breaks both down per locationId.

Limitations

  • Requires a manually managed operator configuration resource. Instances have no Helm-level configuration, so they must be set directly on the Dash0OperatorConfiguration resource. This means the synthetics-worker currently cannot be used together with operator.dash0Export.enabled=true, which has the Helm chart create and manage that resource automatically; install with operator.dash0Export.enabled=false and create the resource yourself instead (see Configuration).
  • Not yet supported on GKE Autopilot. GKE Autopilot requires every container image to be explicitly allowlisted; the synthetics-worker image has not been added to that allowlist yet.