> In my case, all my clients use static ip´s. So,
> I create a inifile with all static ip´s in my LAN
> that can access the database. On login, the client
> app connects to app server via socket and I check
> (on app server) if the remote ip is on list of ip´s
> on inifile. if ok, I allow the client app login in
> database, otherwise I send a message to client app
> to deny entry.

My client uses MS Terminal Server, so there are multiple
users/sessions occuring from the same machine/ip. The same situation
could apply if/when we start using Linux.

The only way I have found to monitor individual user connections is
through the execution of procedures at login and logout. And since
this can leave unwanted details after application failure, this means
a timer based solution for cleanup (or manual) whether such monitoring
is done by a server based service or by the login procedure itself.


Geoff Worboys
Telesis Computing