cerc-io/stack-orchestrator
12
1
Fork 3
Code Issues 117 Pull Requests 25 Actions 1 Packages Projects Releases 493 Wiki Activity

fix: never delete namespace on deployment down

Browse Source 
down() deleted the entire namespace when it wasn't explicitly set in
the spec. This causes a race condition on restart: up() tries to create
resources in a namespace that's still terminating, getting 403 Forbidden.

Always use _delete_resources_by_label() instead. The namespace is cheap
to keep and required for immediate up() after down(). This also matches
the shared-namespace behavior, making down() consistent regardless of
namespace configuration.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This commit is contained in:
A. F. Dudley 2026-03-18 18:47:05 +00:00
parent e298e7444f
commit ae2cea3410
1 changed files with 5 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
stack_orchestrator/deploy/k8s/deploy_k8s.py
View File

@ -581,14 +581,11 @@ class K8sDeployer(Deployer):
if opts.o.debug:
print(f"Error listing PVs: {e}")
# When namespace is explicitly set in the spec, it may be shared with
# other stacks — delete only this stack's resources by label.
# Otherwise the namespace is owned by this deployment, delete it entirely.
shared_namespace = self.deployment_context.spec.get_namespace() is not None
if shared_namespace:
self._delete_resources_by_label(app_label, volumes)
else:
self._delete_namespace()
# Always delete resources by label, never delete the namespace itself.
# Namespace deletion causes a race condition on restart: up() tries to
# create resources in a namespace that's still terminating (403 Forbidden).
# The namespace is cheap to keep and required for immediate up() after down().
self._delete_resources_by_label(app_label, volumes)
if self.is_kind() and not self.skip_cluster_management:
# Destroy the kind cluster