frpc: restart it the first time the admin lines are added (a reload can't work yet)

Found in the VM test: the new install's tunnels never opened, because
frpc only reads admin_addr when it starts, so the first reload failed.
That first time (and whenever a reload fails), frpc is restarted under
systemd and the report says so; later changes reload as before.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
2026-11-17 15:48:28 +02:00
co-authored by Claude Opus 5.5
parent f3f6dc8f7a
commit 19709c2c75
2 changed files with 45 additions and 1 deletions
+24
View File
@@ -103,3 +103,27 @@ EOF
frpc_add hope events 20010 20011
grep -qx 'server_addr = x' "$FRPC_INI"
}
@test "the first time (admin lines just added) frpc is restarted; afterwards it's reloaded" {
cat > "$TMP/bin/systemctl" <<'STUB'
#!/usr/bin/env bash
echo "systemctl $*" >> "$(dirname "$0")/../calls"
STUB
chmod +x "$TMP/bin/systemctl"
PATH="$TMP/bin:$PATH"
FRPC_NEEDS_RESTART=0
frpc_add hope events 20010 20011
[ "$FRPC_NEEDS_RESTART" = 1 ]
run frpc_reload
[ "$status" -eq 0 ]
grep -q "systemctl restart frpc" "$TMP/calls"
[[ "$output" == *"restarting it once"* ]]
# Admin lines now there: the next change reloads.
: > "$TMP/calls"
FRPC_NEEDS_RESTART=0
frpc_add bridge events 20020 20021
[ "$FRPC_NEEDS_RESTART" = 0 ]
frpc_reload
! grep -q "restart frpc" "$TMP/calls"
grep -q reload "$TMP/frp/reloads"
}