From d08d3fb9214ba662e0d9d78dfdcde8cc2412c0cc Mon Sep 17 00:00:00 2001
From: key <let5sne.key@gmail.com>
Date: Wed, 29 Jul 2026 15:25:08 +0800
Subject: [PATCH] fix(release): wait for backend readiness

---
 scripts/release-backend.sh                   | 22 +++++++++++++++++---
 scripts/tests/release-backend-safety.test.sh |  3 +++
 2 files changed, 22 insertions(+), 3 deletions(-)

diff --git a/scripts/release-backend.sh b/scripts/release-backend.sh
index 5759ef4c..2fc3c421 100644
--- a/scripts/release-backend.sh
+++ b/scripts/release-backend.sh
@@ -9,6 +9,7 @@ REMOTE_URL="https://peilian.njzhmj.top"
 REMOTE_BACKUP_ROOT="/opt/wygj/backups"
 PREFLIGHT_SCRIPT="$ROOT_DIR/scripts/release-preflight.sh"
 MIN_FREE_RESERVE_BYTES=$((512 * 1024 * 1024))
+PUBLIC_HEALTH_READY_TIMEOUT_SECONDS=90
 
 fail() {
   echo "release-backend: $*" >&2
@@ -272,6 +273,19 @@ run_public_health() {
     }
 }
 
+wait_for_public_health() {
+  local deadline=$((SECONDS + PUBLIC_HEALTH_READY_TIMEOUT_SECONDS))
+  while ((SECONDS < deadline)); do
+    if run_public_health >/dev/null 2>&1; then
+      echo "release-backend: public endpoints are ready"
+      return 0
+    fi
+    sleep 2
+  done
+  echo "release-backend: public endpoints did not become ready within ${PUBLIC_HEALTH_READY_TIMEOUT_SECONDS}s" >&2
+  run_public_health
+}
+
 run_deploy_plan() {
   validate_local_artifact
   inspect_remote_target
@@ -420,6 +434,8 @@ perform_deploy() {
         restore_remote_backup \
           "$backup_path/ruoyi-admin.jar" "$remote_current_sha256" "$artifact_sha256" \
           || fail "automatic restore failed; use the recorded backup immediately: $backup_path"
+        wait_for_public_health \
+          || fail "candidate activation failed; the original backend was restored but did not become ready"
         fail "candidate activation failed and the original backend was restored"
         ;;
       "$remote_current_sha256")
@@ -431,7 +447,7 @@ perform_deploy() {
     esac
   fi
 
-  if ! run_post_deploy_preflight; then
+  if ! wait_for_public_health || ! run_post_deploy_preflight; then
     after_failure_sha="$(get_remote_target_sha)" \
       || fail "post-deploy preflight failed and the current target hash is unreadable; inspect $backup_path"
     case "$after_failure_sha" in
@@ -440,7 +456,7 @@ perform_deploy() {
         restore_remote_backup \
           "$backup_path/ruoyi-admin.jar" "$remote_current_sha256" "$artifact_sha256" \
           || fail "automatic restore failed; use the recorded backup immediately: $backup_path"
-        run_public_health \
+        wait_for_public_health \
           || fail "original backend was restored but public health verification failed"
         fail "post-deploy preflight failed and the original backend was restored"
         ;;
@@ -552,7 +568,7 @@ REMOTE_ROLLBACK_SAFETY
         ;;
     esac
   fi
-  if ! run_public_health; then
+  if ! wait_for_public_health; then
     after_failure_sha="$(get_remote_target_sha)" \
       || fail "rollback health failed and the current target hash is unreadable; inspect $safety_path"
     case "$after_failure_sha" in
diff --git a/scripts/tests/release-backend-safety.test.sh b/scripts/tests/release-backend-safety.test.sh
index 8da9b52d..318ce445 100644
--- a/scripts/tests/release-backend-safety.test.sh
+++ b/scripts/tests/release-backend-safety.test.sh
@@ -118,6 +118,9 @@ grep -Fq 'require_approval "DEPLOY_BACKEND:$expected_sha256"' "$SCRIPT"
 grep -Fq 'require_approval "ROLLBACK_BACKEND:$expected_sha256"' "$SCRIPT"
 grep -Fq 'deploy and rollback require a clean Git worktree' "$SCRIPT"
 grep -Fq 'run_post_deploy_preflight' "$SCRIPT"
+grep -Fq 'PUBLIC_HEALTH_READY_TIMEOUT_SECONDS=90' "$SCRIPT"
+grep -Fq 'wait_for_public_health' "$SCRIPT"
+grep -Fq 'if ! wait_for_public_health || ! run_post_deploy_preflight; then' "$SCRIPT"
 grep -Fq 'post-deploy preflight failed; restoring' "$SCRIPT"
 grep -Fq 'rollback health failed; restoring safety copy' "$SCRIPT"
 grep -Fq 'expected_target_sha="$5"' "$SCRIPT"