Cannot recover from Bundle Targeting failure due to mismatching hashvalue of existing bundledeployment and secret

[aDisplayName]

Is there an existing issue for this?

• I have searched the existing issues

Current Behavior

Bundle targeting failed with following msg:
terminal error: targeting error: retrying, hash mismatch between secret and bundledeployment
which prevents the update of the bundle from being propagated to BundleDeployments.

Tried to "Force Update" the GitRepo containing the Bundle definition, but the error persists.

Based on the code here:

fleet/internal/cmd/controller/target/builder.go

Line 142 in f32cad3

return nil, fmt.Errorf("retrying, hash mismatch between secret and bundledeployment: actual %s != expected %s", h, bd.Spec.ValuesHash)

It seems the mismatch of the hash in existing bundledeployment, and the option secret of it, preventing the bundle targeting being properly handled. Somehow the mismatch was never got chance to be fixed even "Force Update" is performed on the Bundle object.

Expected Behavior

By force updating the bundle, the bundeldeployment should be updated with the latest content of the Bundle.

Steps To Reproduce

Not Sure.

Environment

- Architecture:
- Fleet Version: v0.14.1
- Cluster:
  - Provider: GKE
  - Options:
  - Kubernetes Version:  v1.33.5-gke.2326000
  - Rancher Server v2.13.0

Logs

{
  "insertId": "cytexg6gd4ppdbwy",
  "jsonPayload": {
    "controller": "bundle",
    "msg": "Reconciler error",
    "controllerGroup": "fleet.cattle.io",
    "ts": "2026-02-09T14:20:17Z",
    "controllerKind": "Bundle",
    "name": "edge-qa-catalog-1-app-catalog-ci",
    "reconcileID": "eda68014-f5ad-4c2c-8477-d843677aaeb2",
    "Bundle": {
      "name": "edge-qa-catalog-1-app-catalog-ci",
      "namespace": "fleet-default"
    },
    "level": "error",
    "stacktrace": "sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller[...]).reconcileHandler\n\t/home/runner/go/pkg/mod/sigs.k8s.io/controller-runtime@v0.22.3/pkg/internal/controller/controller.go:474\nsigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller[...]).processNextWorkItem\n\t/home/runner/go/pkg/mod/sigs.k8s.io/controller-runtime@v0.22.3/pkg/internal/controller/controller.go:421\nsigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller[...]).Start.func1.1\n\t/home/runner/go/pkg/mod/sigs.k8s.io/controller-runtime@v0.22.3/pkg/internal/controller/controller.go:296",
    "error": "terminal error: targeting error: retrying, hash mismatch between secret and bundledeployment: actual 4a2f1ec45890cbdbe041418ef637ebc9b0c53dd3cb0778afcaf3c4bd92269453 != expected be7c99700c2d6b8ef1d90351ff904ab11042fec522bf0ed925c925fb953dd5ff",
    "namespace": "fleet-default"
  },
  "resource": {
    "type": "k8s_container",
    "labels": {
      "location": "europe-west3-b",
      "container_name": "fleet-controller",
      "namespace_name": "cattle-fleet-system",
      "pod_name": "fleet-controller-75bdd597f5-p2bhd"
    }
  },
  "timestamp": "2026-02-09T14:20:17.517678251Z",
  "severity": "ERROR",
  "labels": {
    "logging.gke.io/top_level_controller_name": "fleet-controller",
    "logging.gke.io/top_level_controller_type": "Deployment",
    "k8s-pod/app": "fleet-controller",
    "k8s-pod/pod-template-hash": "75bdd597f5",
    "k8s-pod/fleet_cattle_io/shard-default": "true",
    "compute.googleapis.com/resource_name": "gke-neos-2-wl-ranche-neos-2-wl-ranche-a9064da5-uh6b",
    "k8s-pod/fleet_cattle_io/shard-id": ""
  },
  "logName": "projects/wireline-automation/logs/stderr",
  "receiveTimestamp": "2026-02-09T14:20:22.591335936Z"
}

Anything else?

No response

[aDisplayName]

After we manually delete the Bundle from the rancher, the bundle is then being recreated, and the error then disappeared.

[kkaempf]

Please upgrade to latest Rancher 2.13.2 (Fleet 0.14.2) and report back.

[karlsebal-mcs]

We fixed the issue by setting spec.valuesHash to ''. The bundle will be refreshed and valuesHash recalculated.

Rancher was up-to-date (2.13.2). Removing the bundle works in a way but removes the deployment and redeploys resulting in downtime, which is hardly acceptable for production.

[Wolfgang1966]

Have the same problem here. Rancher 2.13.2, Fleet 0.14.2 Tried setting spec.valuesHash to "". It got recomputed, but the error persists. Also the recomputed hash is neither the actual nor the expected one from the error message.

In my case the affected bundle is longhorn, so deleting it looks like a very bad idea.

Or did you just delete the bundle resource from the rancher cluster without actually uninstalling the bundle itself, @aDisplayName ?

[Wolfgang1966]

Finally I had to delete the bundle which triggered a longhorn uninstall. Luckily longhorn refused being uninstalled as a specific config value has to be explicitly set to allow that, so I deleted the uninstall job in the downstream cluster and the bundle was back in normal state.

Please take care of this issue soon or at least provide a workaround. It has potential to really cause bad trouble.

Thanks a lot!

[kkaempf]

@0xavi0 👆🏻