Idle timeout should be driven by STM for MAC
09-05-2014 04:58 AM
i was going through some user debug logs and saw the following lines:
Sep 3 16:54:49 authmgr: <522246> <DBUG> |authmgr| Idle timeout should be driven by STM for MAC aa:bb:cc:dd:ee:ff.
Sep 3 16:54:29 authmgr: <522247> <DBUG> |authmgr| User idle timer removed for user with MAC aa:bb:cc:dd:ee:ff.
is that normal? if the STM (what does it stand for?) drives the idle timeout, what is the idle time out then?
Re: Idle timeout should be driven by STM for MAC
09-05-2014 08:01 AM - edited 09-05-2014 08:01 AM
These two log entries relate to how we handle users aging out of the system. STM stands for Station Management and it is a control plane process in Controllers and APs. Prior to 6.2, idle timeout was handled by a different process and after that it was handled by STM to improve scalability.
So there are two timers at play here:
- Station timeout period which is a setting under “wlan ssid-profile <name>”. The default is 1000 seconds.
- Idle timeout period is a setting under “aaa timer”. The default is 300 seconds.
In 6.4.1 we also have some aaa-profile timers to override the global aaa but let's put that aside for this discussion.
There are two predominate situations where these come into play.
Client sends a disassociate from an AP.
--> In this situation, once the client dis-associates, the user will age out according to the aaa idle timer.
Client doesn't send a disassociate but stops sending traffic (e.g. walks away)
--> In this situation, the client will age out according to the station timeout.
Now going to your specific log entries, these are typically seen during user creation. I suspect if we had the full log set, we would see a "Station UP" event preceding these messages. Is that the case?