Drift correction not working

[lindhe]

Is there an existing issue for this?

• I have searched the existing issues

Current Behavior

If an object is changed, Fleet detects the diff but does nothing to converge to a healthy state.

Expected Behavior

When spec.correctDrift.enabled=true, I expect Fleet to try and apply changes as soon as there is a diff.

Steps To Reproduce

1. Have Rancher v2.8.1 installed.

2. In Rancher, click "Continuous Delivery" and "Git Repos" and select the "fleet-local" workspace.

3. Add a GitRepo that applies some resource. Make sure to check "Enable Self-Healing" to set spec.correctDrift.enabled=true in the bundle.

4. Wait for the GitRepo to sync and become healthy, with the new resource created and in state "Ready".

5. Edit the resource using kubectl edit, e.g. delete a label or something.

6. Observe new state "Modified" for the resource:

   

Environment

- Architecture: amd64
- Fleet Version: The one that's bundled with Rancher v2.8.1. 
- Cluster:
  - Provider: RKE2
  - Options: 3 nodes upstream cluster
  - Kubernetes Version: 1.27.9

Logs

No response

Anything else?

It looks like #1594 tried to implement drift correction, but it's clearly not working.

[manno]

Probably related to #2551

[jhoblitt]

I'm seeing the same behavior with rancher 2.8.5 / fleet 0.9.8.

[jhoblitt]

I've reproduced the problem with rancher 2.9.1 / 0.10.1 as well:

The fleet-agent logs on the cluster are the same messages repeated over and over again. E.g.:

{"level":"info","ts":"2024-09-04T22:15:40Z","logger":"bundledeployment.RemoveExternalChanges","msg":"Drift correction: rollback","controller":"bundledeployment","controllerGroup":"fleet.cattle.io","controllerKind":"BundleDeployment","BundleDeployment":{"name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb"},"namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb","name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","reconcileID":"c6f34409-8c39-4bfa-bb86-ff79d3028f46"}
{"level":"info","ts":"2024-09-04T22:15:41Z","logger":"bundledeployment.DeployBundle","msg":"Deployed bundle","controller":"bundledeployment","controllerGroup":"fleet.cattle.io","controllerKind":"BundleDeployment","BundleDeployment":{"name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb"},"namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb","name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","reconcileID":"ae91a7c9-9d4e-4ed4-988a-eeb61c5d015d","deploymentID":"s-f4e58f4e8d63737718ef1c935b3bcd8054daea4ce155c1be7f814b4738481:704080e856144689404f7488b30ba700c635fea680a58494dd909c76b226262d","appliedDeploymentID":"s-f4e58f4e8d63737718ef1c935b3bcd8054daea4ce155c1be7f814b4738481:704080e856144689404f7488b30ba700c635fea680a58494dd909c76b226262d","release":"rook-ceph/rook-ceph-conf:1033","appliedDeploymentID":"s-f4e58f4e8d63737718ef1c935b3bcd8054daea4ce155c1be7f814b4738481:704080e856144689404f7488b30ba700c635fea680a58494dd909c76b226262d"}
{"level":"info","ts":"2024-09-04T22:15:42Z","logger":"bundledeployment.UpdateStatus","msg":"Status not ready","controller":"bundledeployment","controllerGroup":"fleet.cattle.io","controllerKind":"BundleDeployment","BundleDeployment":{"name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb"},"namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb","name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","reconcileID":"ae91a7c9-9d4e-4ed4-988a-eeb61c5d015d","error":"cephnfs.ceph.rook.io rook-ceph/auxtel modified {\"spec\":{\"server\":{\"resources\":{\"limits\":{\"cpu\":\"3\"}}}}}"}
{"level":"info","ts":"2024-09-04T22:15:42Z","logger":"bundledeployment.RemoveExternalChanges","msg":"Drift correction: rollback","controller":"bundledeployment","controllerGroup":"fleet.cattle.io","controllerKind":"BundleDeployment","BundleDeployment":{"name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb"},"namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb","name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","reconcileID":"ae91a7c9-9d4e-4ed4-988a-eeb61c5d015d"}
{"level":"info","ts":"2024-09-04T22:15:43Z","logger":"bundledeployment.DeployBundle","msg":"Deployed bundle","controller":"bundledeployment","controllerGroup":"fleet.cattle.io","controllerKind":"BundleDeployment","BundleDeployment":{"name":"ruka-fleet-s-dev-c-ruka-rook-ceph-cluster","namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb"},"namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb","name":"ruka-fleet-s-dev-c-ruka-rook-ceph-cluster","reconcileID":"c1678532-73f5-4929-9985-e50db5603133","deploymentID":"s-1c47fdd307de7cd53771b1bcf05e2d5bf014de495953153ac73876102439a:096621fc89c2b75c31148d0b150ad30027e7035383a85288a66c67d0980fb055","appliedDeploymentID":"s-1c47fdd307de7cd53771b1bcf05e2d5bf014de495953153ac73876102439a:096621fc89c2b75c31148d0b150ad30027e7035383a85288a66c67d0980fb055","release":"rook-ceph/rook-ceph-cluster:2","appliedDeploymentID":"s-1c47fdd307de7cd53771b1bcf05e2d5bf014de495953153ac73876102439a:096621fc89c2b75c31148d0b150ad30027e7035383a85288a66c67d0980fb055"}
{"level":"info","ts":"2024-09-04T22:15:43Z","logger":"bundledeployment.DeployBundle","msg":"Deployed bundle","controller":"bundledeployment","controllerGroup":"fleet.cattle.io","controllerKind":"BundleDeployment","BundleDeployment":{"name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb"},"namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb","name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","reconcileID":"1407fd06-6c0a-4e3a-98a8-baf4bff370da","deploymentID":"s-f4e58f4e8d63737718ef1c935b3bcd8054daea4ce155c1be7f814b4738481:704080e856144689404f7488b30ba700c635fea680a58494dd909c76b226262d","appliedDeploymentID":"s-f4e58f4e8d63737718ef1c935b3bcd8054daea4ce155c1be7f814b4738481:704080e856144689404f7488b30ba700c635fea680a58494dd909c76b226262d","release":"rook-ceph/rook-ceph-conf:1034","appliedDeploymentID":"s-f4e58f4e8d63737718ef1c935b3bcd8054daea4ce155c1be7f814b4738481:704080e856144689404f7488b30ba700c635fea680a58494dd909c76b226262d"}
{"level":"info","ts":"2024-09-04T22:15:44Z","logger":"bundledeployment.UpdateStatus","msg":"Status not ready","controller":"bundledeployment","controllerGroup":"fleet.cattle.io","controllerKind":"BundleDeployment","BundleDeployment":{"name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb"},"namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb","name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","reconcileID":"1407fd06-6c0a-4e3a-98a8-baf4bff370da","error":"cephnfs.ceph.rook.io rook-ceph/auxtel modified {\"spec\":{\"server\":{\"resources\":{\"limits\":{\"cpu\":\"3\"}}}}}"}
{"level":"info","ts":"2024-09-04T22:15:44Z","logger":"bundledeployment.RemoveExternalChanges","msg":"Drift correction: rollback","controller":"bundledeployment","controllerGroup":"fleet.cattle.io","controllerKind":"BundleDeployment","BundleDeployment":{"name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb"},"namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb","name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","reconcileID":"1407fd06-6c0a-4e3a-98a8-baf4bff370da"}
{"level":"info","ts":"2024-09-04T22:15:45Z","logger":"bundledeployment.DeployBundle","msg":"Deployed bundle","controller":"bundledeployment","controllerGroup":"fleet.cattle.io","controllerKind":"BundleDeployment","BundleDeployment":{"name":"ruka-fleet-s-dev-c-ruka-rook-ceph-cluster","namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb"},"namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb","name":"ruka-fleet-s-dev-c-ruka-rook-ceph-cluster","reconcileID":"494aa1f3-8e09-4790-ac4a-7acc6bc34b7f","deploymentID":"s-1c47fdd307de7cd53771b1bcf05e2d5bf014de495953153ac73876102439a:096621fc89c2b75c31148d0b150ad30027e7035383a85288a66c67d0980fb055","appliedDeploymentID":"s-1c47fdd307de7cd53771b1bcf05e2d5bf014de495953153ac73876102439a:096621fc89c2b75c31148d0b150ad30027e7035383a85288a66c67d0980fb055","release":"rook-ceph/rook-ceph-cluster:2","appliedDeploymentID":"s-1c47fdd307de7cd53771b1bcf05e2d5bf014de495953153ac73876102439a:096621fc89c2b75c31148d0b150ad30027e7035383a85288a66c67d0980fb055"}
{"level":"info","ts":"2024-09-04T22:15:45Z","logger":"bundledeployment.DeployBundle","msg":"Deployed bundle","controller":"bundledeployment","controllerGroup":"fleet.cattle.io","controllerKind":"BundleDeployment","BundleDeployment":{"name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb"},"namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb","name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","reconcileID":"3f0b30ce-a868-48e0-95aa-1ce063f78d48","deploymentID":"s-f4e58f4e8d63737718ef1c935b3bcd8054daea4ce155c1be7f814b4738481:704080e856144689404f7488b30ba700c635fea680a58494dd909c76b226262d","appliedDeploymentID":"s-f4e58f4e8d63737718ef1c935b3bcd8054daea4ce155c1be7f814b4738481:704080e856144689404f7488b30ba700c635fea680a58494dd909c76b226262d","release":"rook-ceph/rook-ceph-conf:1035","appliedDeploymentID":"s-f4e58f4e8d63737718ef1c935b3bcd8054daea4ce155c1be7f814b4738481:704080e856144689404f7488b30ba700c635fea680a58494dd909c76b226262d"}
{"level":"info","ts":"2024-09-04T22:15:46Z","logger":"bundledeployment.UpdateStatus","msg":"Status not ready","controller":"bundledeployment","controllerGroup":"fleet.cattle.io","controllerKind":"BundleDeployment","BundleDeployment":{"name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb"},"namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb","name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","reconcileID":"3f0b30ce-a868-48e0-95aa-1ce063f78d48","error":"cephnfs.ceph.rook.io rook-ceph/auxtel modified {\"spec\":{\"server\":{\"resources\":{\"limits\":{\"cpu\":\"3\"}}}}}"}
{"level":"info","ts":"2024-09-04T22:15:46Z","logger":"bundledeployment.RemoveExternalChanges","msg":"Drift correction: rollback","controller":"bundledeployment","controllerGroup":"fleet.cattle.io","controllerKind":"BundleDeployment","BundleDeployment":{"name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb"},"namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb","name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","reconcileID":"3f0b30ce-a868-48e0-95aa-1ce063f78d48"}
{"level":"info","ts":"2024-09-04T22:15:47Z","logger":"bundledeployment.DeployBundle","msg":"Deployed bundle","controller":"bundledeployment","controllerGroup":"fleet.cattle.io","controllerKind":"BundleDeployment","BundleDeployment":{"name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb"},"namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb","name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","reconcileID":"9cd5479d-4754-4c7b-909f-3acde128815b","deploymentID":"s-f4e58f4e8d63737718ef1c935b3bcd8054daea4ce155c1be7f814b4738481:704080e856144689404f7488b30ba700c635fea680a58494dd909c76b226262d","appliedDeploymentID":"s-f4e58f4e8d63737718ef1c935b3bcd8054daea4ce155c1be7f814b4738481:704080e856144689404f7488b30ba700c635fea680a58494dd909c76b226262d","release":"rook-ceph/rook-ceph-conf:1036","appliedDeploymentID":"s-f4e58f4e8d63737718ef1c935b3bcd8054daea4ce155c1be7f814b4738481:704080e856144689404f7488b30ba700c635fea680a58494dd909c76b226262d"}
{"level":"info","ts":"2024-09-04T22:15:48Z","logger":"bundledeployment.UpdateStatus","msg":"Status not ready","controller":"bundledeployment","controllerGroup":"fleet.cattle.io","controllerKind":"BundleDeployment","BundleDeployment":{"name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb"},"namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb","name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","reconcileID":"9cd5479d-4754-4c7b-909f-3acde128815b","error":"cephnfs.ceph.rook.io rook-ceph/auxtel modified {\"spec\":{\"server\":{\"resources\":{\"limits\":{\"cpu\":\"3\"}}}}}"}
{"level":"info","ts":"2024-09-04T22:15:48Z","logger":"bundledeployment.RemoveExternalChanges","msg":"Drift correction: rollback","controller":"bundledeployment","controllerGroup":"fleet.cattle.io","controllerKind":"BundleDeployment","BundleDeployment":{"name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb"},"namespace":"cluster-it-5628-ruka-plus-plus-ruka-67e02fc747cb","name":"ruka-fleet-s-dev-c-ruka-rook-ceph-conf","reconcileID":"9cd5479d-4754-4c7b-909f-3acde128815b"}

[weyfonk]

I have tried, and failed, to reproduce this against the current main by deleting a label on a config map. This needs further investigation.
Could you share an example of a workload (GitRepo), or a known manifest or chart, which triggers this failure?

In any case, #2917 should reduce the noise compared to logs shared above.