ci: add VPN diagnostics and curl retry to server-tests workflow

The "Fetch SASJS server" step was timing out after 130s+ despite the
VPN reporting "Connected", with no way to tell whether routes hadn't
propagated yet or the firewall was silently dropping packets from the
runner's IP. Log VPN session/route/DNS state right after connecting,
and retry the curl with a shorter per-attempt timeout so a routing
race resolves itself instead of hanging the whole job.

.github/workflows/server-tests.yml

@@ -86,8 +86,24 @@ jobs:
       - name: install pm2
         run: npm i -g pm2
 
+      - name: Diagnose VPN connectivity
+        run: |
+          echo "--- openvpn3 sessions-list ---"
+          openvpn3 sessions-list
+          echo "--- openvpn3 session-stats ---"
+          openvpn3 session-stats --config .github/vpn/config.ovpn
+          echo "--- ip addr (tun interface) ---"
+          ip addr show tun0 || echo "no tun0 interface found"
+          echo "--- ip route ---"
+          ip route
+          echo "--- DNS resolution for target host ---"
+          getent hosts "$(echo '${{ secrets.SASJS_SERVER_URL }}' | sed -E 's#^[a-z]+://##; s#[:/].*##')" || echo "DNS resolution failed"
+        shell: bash
+
       - name: Fetch SASJS server
-        run: curl ${{ secrets.SASJS_SERVER_URL }}/SASjsApi/info
+        run: |
+          curl -v --retry 5 --retry-delay 5 --retry-all-errors --connect-timeout 15 --max-time 30 \
+            "${{ secrets.SASJS_SERVER_URL }}/SASjsApi/info"
 
       - name: Deploy sasjs-tests
         run: |