this limit of simultaneous winbox/dude connections is generating a very big problem in our network,

as our network is a little big, each region has its router and sectors mapped by the dude and I select that the device is RouterOS,
with this I manage to monitor the signals of wireless clients and support them in a single point from our CCR/DUDE in /dude ros registration-table because there are many clients

however, the RouterOS devices monitored (routers of neighborhoods and sectors) do not stop reconnecting connected/disconnected via dude even though it is a fiber link between regions, and with that it is very difficult to support customers, all customers disappear and reappear all the time. CCR wireless registration (/dude ros registration-table) among many other parameters
and here we also use alarms via telegram when the signal quality ccq_nv2 dropped foreach ospf neighbor etc via "ros_command", among many other parameters that is only possible via RouterOS and which does not have via snmp,
and now many of the ccq alerts and other parameters via telegram don't stop coming because the connections / disconnections from dude to the other monitored RouterOS devices don't stop happening,


this is not a problem reported only with me, but with many other big RouterBoards consumers: viewtopic.php?t=160451
the last stable version of the server that works normally is: 6.45.9
sometimes I can't access the CCR either via Winbox or via API

please return to having unlimited winbox/dude/ros_command connections, so the only limitation would be the CPU of the CCR/CHR/DUDE

Do you only have one of The Dude?
Do you have all CPEs / Users inside The Dude ???
Didn't you just put the "transport and AP" network?
Do you have a single "point of failure" instead of distributing the various things between CPE and AP, and other devices?
For example, why is it not the AP, where the Station is registered, sending the alert?