stack/monitor: Fix auto-restart (used bad variable name)

/reviewed-on nexedi/slapos!500

stack/monitor/buildout.hash.cfg

@@ -14,4 +14,4 @@
 # not need these here).
 [monitor2-template]
 filename = instance-monitor.cfg.jinja2.in
-md5sum = 3b3acb2291fc7458bb11efc80a5aba27
+md5sum = 74bc641e97aa07a0cb1fb442c5ede183

stack/monitor/instance-monitor.cfg.jinja2.in

@@ -410,7 +410,7 @@ depends =
   ${monitor-collect-cron-entry:name}
   ${cron-entry-logrotate:name}
   ${logrotate-entry-cron:name}
-  ${certificate-authority-service:wrapper}
+  ${certificate-authority-service:wrapper-path}
   ${monitor-conf:rendered}
   ${start-monitor:wrapper-path}
   ${ca-monitor-httpd-service:wrapper-path}