From 637bdc8f933f832c55dae3b32597829f6bf79871 Mon Sep 17 00:00:00 2001 From: Bryan Cox Date: Mon, 3 Aug 2026 10:49:21 -0400 Subject: [PATCH] feat(hypershift-azure-deprovision): add --force flag and reduce grace period Integrates the --force flag from openshift/hypershift#9134 into the periodic Azure deprovisioner job. When the 20m grace period expires, --force strips all finalizers from child resources bottom-up and continues with best-effort infrastructure cleanup. Reduces --cluster-grace-period from 40m to 20m. With 27+ stuck HCs at the previous 40m/cluster, Phase 1 exhausted the 4h Prow budget. At 20m/cluster with --force, stuck HCs are force-cleaned within the budget. Fixes CNTRLPLANE-3943 Co-Authored-By: Claude Sonnet 4.5 --- .../azure/deprovision/hypershift-azure-deprovision-commands.sh | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/ci-operator/step-registry/hypershift/azure/deprovision/hypershift-azure-deprovision-commands.sh b/ci-operator/step-registry/hypershift/azure/deprovision/hypershift-azure-deprovision-commands.sh index c08510adf12e7..9ce24d8d1c583 100755 --- a/ci-operator/step-registry/hypershift/azure/deprovision/hypershift-azure-deprovision-commands.sh +++ b/ci-operator/step-registry/hypershift/azure/deprovision/hypershift-azure-deprovision-commands.sh @@ -82,7 +82,8 @@ phase1_hc_pruner() { --name="${hc}" \ --namespace=clusters \ --dns-zone-rg-name=os4-common \ - --cluster-grace-period 40m || had_failure=$((had_failure+1)) + --cluster-grace-period 20m \ + --force || had_failure=$((had_failure+1)) fi done