Diffing Customization

It is possible for an application to be OutOfSync even immediately after a successful Sync operation. Some reasons for this might be:

  • There is a bug in the manifest, where it contains extra/unknown fields from the actual K8s spec. These extra fields would get dropped when querying Kubernetes for the live state, resulting in an OutOfSync status indicating a missing field was detected.
  • The sync was performed (with pruning disabled), and there are resources which need to be deleted.
  • A controller or mutating webhook is altering the object after it was submitted to Kubernetes in a manner which contradicts Git.
  • A Helm chart is using a template function such as randAlphaNum, which generates different data every time helm template is invoked.
  • For Horizontal Pod Autoscaling (HPA) objects, the HPA controller is known to reorder spec.metrics in a specific order. See kubernetes issue #74099. To work around this, you can order spec.metrics in Git in the same order that the controller prefers.

In case it is impossible to fix the upstream issue, Argo CD allows you to optionally ignore differences of problematic resources. The diffing customization can be configured for single or multiple application resources or at a system level.

Application Level Configuration

Argo CD allows ignoring differences at a specific JSON path, using RFC6902 JSON patches and JQ path expressions). It is also possible to ignore differences from fields owned by specific managers defined in metadata.managedFields in live resources.

The following sample application is configured to ignore differences in spec.replicas for all deployments:

  1. spec:
  2.   ignoreDifferences:
  3.   - group: apps
  4.     kind: Deployment
  5.     jsonPointers:
  6.     - /spec/replicas

Note that the group field relates to the Kubernetes API group without the version. The above customization could be narrowed to a resource with the specified name and optional namespace:

  1. spec:
  2.   ignoreDifferences:
  3.   - group: apps
  4.     kind: Deployment
  5.     name: guestbook
  6.     namespace: default
  7.     jsonPointers:
  8.     - /spec/replicas

To ignore elements of a list, you can use JQ path expressions to identify list items based on item content:

  1. spec:
  2.   ignoreDifferences:
  3.   - group: apps
  4.     kind: Deployment
  5.     jqPathExpressions:
  6.     - .spec.template.spec.initContainers[] | select(.name == "injected-init-container")

To ignore fields owned by specific managers defined in your live resources:

  1. spec:
  2.   ignoreDifferences:
  3.   - group: "*"
  4.     kind: "*"
  5.     managedFieldsManagers:
  6.     - kube-controller-manager

The above configuration will ignore differences from all fields owned by kube-controller-manager for all resources belonging to this application.

If you have a slash / in your pointer path, you can use the ~1 character. For example:

  1. spec:
  2.   ignoreDifferences:
  3.   - kind: Node
  4.     jsonPointers: /metadata/labels/node-role.kubernetes.io~1worker

System-Level Configuration

The comparison of resources with well-known issues can be customized at a system level. Ignored differences can be configured for a specified group and kind in resource.customizations key of argocd-cm ConfigMap. Following is an example of a customization which ignores the caBundle field of a MutatingWebhookConfiguration webhooks:

  1. data:
  2.   resource.customizations.ignoreDifferences.admissionregistration.k8s.io_MutatingWebhookConfiguration: |
  3.     jqPathExpressions:
  4.     - '.webhooks[]?.clientConfig.caBundle'

Resource customization can also be configured to ignore all differences made by a managedField.manager at the system level. The example below shows how to configure Argo CD to ignore changes made by kube-controller-manager in Deployment resources.

  1. data:
  2.   resource.customizations.ignoreDifferences.apps_Deployment: |
  3.     managedFieldsManagers:
  4.     - kube-controller-manager

It is possible to configure ignoreDifferences to be applied to all resources in every Application managed by an Argo CD instance. In order to do so, resource customizations can be configured like in the example below:

  1. data:
  2.   resource.customizations.ignoreDifferences.all: |
  3.     managedFieldsManagers:
  4.     - kube-controller-manager
  5.     jsonPointers:
  6.     - /spec/replicas

The status field of CustomResourceDefinitions is often stored in Git/Helm manifest and should be ignored during diffing. The ignoreResourceStatusField setting simplifies handling that edge case:

  1. data:
  2.   resource.compareoptions: |
  3.     # disables status field diffing in specified resource types
  4.     # 'crd' - CustomResourceDefinitions (default)
  5.     # 'all' - all resources
  6.     # 'none' - disabled
  7.     ignoreResourceStatusField: crd

By default status field is ignored during diffing for CustomResourceDefinition resource. The behavior can be extended to all resources using all value or disabled using none.

Ignoring RBAC changes made by AggregateRoles

If you are using Aggregated ClusterRoles and don’t want Argo CD to detect the rules changes as drift, you can set resource.compareoptions.ignoreAggregatedRoles: true. Then Argo CD will no longer detect these changes as an event that requires syncing.

  1. apiVersion: v1
  2. kind: ConfigMap
  3. metadata:
  4.   name: argocd-cm
  5. data:
  6.   resource.compareoptions: |
  7.     # disables status field diffing in specified resource types
  8.     ignoreAggregatedRoles: true

Known Kubernetes types in CRDs (Resource limits, Volume mounts etc)

Some CRDs are re-using data structures defined in the Kubernetes source base and therefore inheriting custom JSON/YAML marshaling. Custom marshalers might serialize CRDs in a slightly different format that causes false positives during drift detection.

A typical example is the argoproj.io/Rollout CRD that re-using core/v1/PodSpec data structure. Pod resource requests might be reformatted by the custom marshaller of IntOrString data type:

from:

  1. resources:
  2.   requests:
  3.     cpu: 100m

to:

  1. resources:
  2.   requests:
  3.     cpu: 0.1

The solution is to specify which CRDs fields are using built-in Kubernetes types in the resource.customizations section of argocd-cm ConfigMap:

  1. apiVersion: v1
  2. kind: ConfigMap
  3. metadata:
  4.   name: argocd-cm
  5.   namespace: argocd
  6.   labels:
  7.     app.kubernetes.io/name: argocd-cm
  8.     app.kubernetes.io/part-of: argocd
  9. data:
  10.   resource.customizations.knownTypeFields.argoproj.io_Rollout: |
  11.     - field: spec.template.spec
  12.       type: core/v1/PodSpec

The list of supported Kubernetes types is available in diffing_known_types.txt

JQ Path expression timeout

By default, the evaluation of a JQPathExpression is limited to one second. If you encounter a “JQ patch execution timed out” error message due to a complex JQPathExpression that requires more time to evaluate, you can extend the timeout period by configuring the ignore.normalizer.jq.timeout setting within the argocd-cmd-params-cm ConfigMap.

```yaml apiVersion: v1 kind: ConfigMap metadata: name: argocd-cmd-params-cm data: ignore.normalizer.jq.timeout: “5s”