Hi all,
I’m trying to figure out what method the Mikrotik “keepalive timeout” uses to determine whether the client is still connected or not …
I found the question asked a few times already in the forum, but no authoratative answer was given.
Does keepalive timeout rely on some sort of method (ie icmp ping) that a client firewall (ie: xp sp2 firewall) can block? or does it use some other method, such as arp requests/replies?
Thanks mucho,
-Rob
Can anyone answer this? I, myself, have previously asked this question, but I never received a response…Thanks
From the manual under HotSpot Gateway, HotSpot Active Users:
keepalive-lost (read-only: time) - how much time past since last packed [sic] from the client has been received
cybertime, that sounds more like an idle timeout to me…
it could be possible (but maybe unlikely) that a client is still connected for longer than the keepalive-timeout and simply not sending any packets.
my goal in using keepalive-timeout rather than idle-timeout is to avoid logging off idle (but still present) customers…
-rob
keep-alive timeouts are used to determine weather or not a session is still active. When a client isn’t sending any data it will send out keep-alive messages to the hotspot/router, if they exceed the keep-alive timer the session is declared failed, and torn down.
It’s not ping based as we have some public Hotspots that use the keepalive and I know some of the users have heavy firewalls set. Once people disconnect or shutdown their computer, the keepalive kicks in and they get disconnected shortly after. You can be idle for quite some time and still stay connected - as long as the client has a good wireless connection. It works as advertised…
To add to what jarosoup said, keep alives are outbound from client to router/server so unless the user is using outbound firewall rules the firewall shouldn’t matter.