ci: run app directly with health check polling before systemd restart

Runs node dist/main.js in background, polls localhost:3000/api for up
to 30s. If app crashes, prints startup log. This definitively
identifies whether it's a code error or systemd timeout issue.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

.gitea/workflows/deploy.yml

@@ -29,7 +29,6 @@ jobs:
           cd /tmp/api-server
           npx prisma generate
           npm run build
-          # Normalize output: ensure main.js is at dist/main.js
           if [ ! -f dist/main.js ] && [ -f dist/src/main.js ]; then
             echo "[build] Moving dist/src/* to dist/"
             mv dist/src/* dist/
@@ -86,20 +85,41 @@ jobs:
         run: |
           cd /opt/zhixi/backend && npx prisma generate
 
-      - name: Test-run API directly to capture errors
+      - name: Test-run and health check
         run: |
+          # Stop systemd service first to free the port
+          sudo systemctl stop zhixi-api 2>/dev/null || true
+          sleep 1
+          # Run app directly in background
           cd /opt/zhixi/backend
-          timeout 10 node dist/main.js 2>&1 || true
+          PORT=3000 node dist/main.js > /tmp/zhixi-startup.log 2>&1 &
+          APP_PID=$!
+          echo "[deploy] App PID: $APP_PID"
+          # Wait for app to start (up to 30s)
+          for i in $(seq 1 30); do
+            sleep 1
+            if curl -sf http://localhost:3000/api > /dev/null 2>&1; then
+              echo "[deploy] API healthy after ${i}s!"
+              break
+            fi
+            if ! kill -0 $APP_PID 2>/dev/null; then
+              echo "[deploy] App crashed after ${i}s — startup log:"
+              cat /tmp/zhixi-startup.log
+              exit 1
+            fi
+          done
+          # Kill test instance
+          kill $APP_PID 2>/dev/null
+          wait $APP_PID 2>/dev/null
 
       - name: Restart API service
         run: |
           sudo systemctl restart zhixi-api
-          sleep 3
+          sleep 5
           if sudo systemctl is-active zhixi-api; then
             echo "[deploy] zhixi-api active OK"
           else
-            echo "[deploy] zhixi-api FAILED to start — checking logs:"
-            sudo journalctl -u zhixi-api --no-pager -n 100
+            echo "[deploy] zhixi-api FAILED to start"
             exit 1
           fi