Pre and Post Delete CHecks
Sveltos provides a robust framework to ensure your managed clusters are in the correct state throughout the entire lifecycle of a resource—from pre-deployment gating to final deletion. This is achieved through four types of checks: PreDeployChecks, ValidateHealthChecks, PreDeleteChecks, and PostDeleteChecks.
These checks are defined within the spec section of a ClusterProfile or a Profile.
Pre-Deploy Checks
PreDeployChecks are executed before Sveltos deploys any resources for a given feature. If any check fails, deployment is held and retried later — no resources are applied until all checks pass. This is useful for enforcing prerequisites that must exist in the target cluster before an application is rolled out.
spec:
preDeployChecks:
- name: "require-storage-class"
featureID: Resources
group: "storage.k8s.io"
version: "v1"
kind: "StorageClass"
labelFilters:
- key: tier
operation: Equal
value: ssd
In the example above, Sveltos will not deploy the Resources feature until at least one StorageClass with the label tier=ssd exists in the managed cluster.
Validate Health Checks
ValidateHealthChecks are executed after resources are deployed. They ensure that the application or infrastructure is not just present, but actually functional.
For more details on how to configure these during deployment, see Depends On with Health Checks.
Pre-Delete Checks
PreDeleteChecks are executed before Sveltos begins the deletion process. If any of these checks fail, Sveltos will halt the deletion. This is essential for ensuring that prerequisites (like data backups or draining connections) are met before infrastructure is torn down.
spec:
preDeleteChecks:
- name: "check-no-active-pods"
featureID: Resources
group: ""
version: "v1"
kind: "Pod"
namespace: "production"
script: |
function evaluate()
-- If any pods are found, we return health=false to stop deletion
return { health = false, message = "Active pods still exist in production" }
end
Post-Delete Checks
PostDeleteChecks are executed after Sveltos has issued the delete commands for all resources. These checks ensure the environment has reached a "clean" state, verifying that no lingering resources (like orphaned LoadBalancers or PVCs) remain.