hull reconcile

hull reconcile re-applies a release’s stored manifest onto the cluster, pulling drifted resources back to the state hull last recorded.

When to use it

After hull drift reports cluster drift — someone ran kubectl edit, a controller patched a field, or a resource was deleted out of band — and you want the cluster back to the recorded state.
As a converge step in automation that keeps live resources matching hull’s stored state, without cutting a new revision.

What happens

Reads the latest stored manifest for <release-name>.
Compares it against the live cluster; if nothing has drifted, it prints No drift to reconcile. and stops.
Otherwise it re-applies the stored resources, skipping any annotated resource-policy: keep (their drift is intentionally preserved).
Unless --no-wait is set, it waits up to --timeout for the re-applied resources to become Ready.
Prints how many resources it converged and lists them.

Mutating: it writes to the cluster. It does not create a new revision — the package and values are unchanged, so use hull upgrade when you actually want to roll new content. Requires a reachable cluster.

Usage

hull reconcile <release-name> [flags]

Flags

Flag	Type	Default	Description
`--timeout`	duration	5m0s	how long to wait for re-applied resources to become Ready
`--no-wait`	—	false	return as soon as the manifest is applied, without waiting for readiness

Persistent flags inherited from `hull`

Flag	Type	Description
`--debug`	—	enable debug output
`--kube-context`	string	Kubernetes context to use
`--kubeconfig`	string	path to kubeconfig file
`-n, --namespace`	string	Kubernetes namespace

Worked example — push stored state back onto a drifted cluster

INPUT — the cluster has drifted. Someone edited the live Service port, and hull drift flags it:

hull drift mychart -n apps
# ~ differs   Service/mychart  (namespace apps)
#       spec.ports.0.port  ⚠ cluster drift
#           state:   8080
#           running: 9090

The stored state says 8080; the live cluster says 9090.

Reconcile the release:

hull reconcile mychart -n apps

OUTPUT:

Reconciled 1 resource(s):
  - Service/mychart

Re-run drift to confirm it is clean:

hull drift mychart -n apps
# 0 cluster-drift, 0 pending-apply, 0 orphan, 0 missing, 0 to-create.

Tracing the output:

Output	Cause
`Reconciled 1 resource(s)`	one resource had drifted from the stored state
`- Service/mychart`	the Service whose live port `9090` was re-applied back to the stored `8080`
drift now reports `0 cluster-drift`	the cluster matches the stored manifest again

If nothing had drifted, reconcile would instead print No drift to reconcile. and leave the cluster alone.