From 0f37945f51942ee64d1e059937420b1d2b049ce5 Mon Sep 17 00:00:00 2001 From: "juvenal.diaz" Date: Sat, 26 Sep 2026 09:19:46 -0600 Subject: [PATCH] tolerate transient k8s api fails --- bootstrap/cluster/main.tf | 16 ++++++++++++---- 1 file changed, 12 insertions(+), 4 deletions(-) diff --git a/bootstrap/cluster/main.tf b/bootstrap/cluster/main.tf index 34ad772..cdd491d 100644 --- a/bootstrap/cluster/main.tf +++ b/bootstrap/cluster/main.tf @@ -445,7 +445,7 @@ resource "null_resource" "kube_system_resource_policy" { triggers = { kubeconfig_path = var.kubeconfig_path - resource_policy_version = "2" + resource_policy_version = "3" } provisioner "local-exec" { @@ -508,11 +508,19 @@ wait_for_apiserver_ready() { } wait_for_apiserver_stable() { - wait_for_apiserver_ready - for _ in $(seq 1 6); do + stable_checks=0 + for _ in $(seq 1 120); do + if kubectl --kubeconfig "${self.triggers.kubeconfig_path}" get --raw=/readyz >/dev/null 2>&1; then + stable_checks=$((stable_checks + 1)) + if [ "$stable_checks" -ge 6 ]; then + return 0 + fi + else + stable_checks=0 + fi sleep 5 - kubectl --kubeconfig "${self.triggers.kubeconfig_path}" get --raw=/readyz >/dev/null done + kubectl --kubeconfig "${self.triggers.kubeconfig_path}" get --raw=/readyz >/dev/null } patch_static_pod_resources etcd 100m 256Mi 1000m 2Gi