So, a god strategy can be:
stop al cron
* * * * * /usr/local/ispconfig/server/server.sh > /dev/null 2>> /var/log/ispconfig/cron.log
30 00 * * * /usr/local/ispconfig/server/cron_daily.sh > /dev/null 2>> /var/log/ispconfig/cron.log
an all nodes. Then, upgrade master, then for each node i upgrade, I restart cron.
Thi sway I haven't any polling by nodes that query a db structure different by that they expect and all is clean and clear