We're running AREV in and NT4.0 SP5 environment with Win 95 FAT16 on the desktops. We're also running users in across the network, via a Cubix, both network and dialin.
This problem begins at a remote location. They go into a file and when they go to save their changes, because an existing link has maxed out it's storage, it locks up. Sometimes it will let them go into debugger, usually you can get out by hitting escape and sometimes they just have to CTL,ATL,DEL to get out of AREV. The end result is a lock on the record that will not go away until you bounce the server.
Any alternatives to bouncing the server? It's a pain to get 150+ users in three different locations off just to bounce it. There's got to be a easier way . . .
Do all of your remote users use distinct logins, or is it a shared login name as far as the server is concerned?
If they are distinct logins, having the user log out and remain out for 5 - 10 minutes should allow the server time enough to time-out the session and release any locks.
Matt,
They're all unique - and we'd never considered leaving them off for that length of time. Obviously having remote users is relatively new to us. In house, simply rebooting the PC has always taken care of any locks.
15 minutes appears to be the optimum amount of time for us with our remote users. THANK YOU!
If I'd have taken down the server and/or AREV one more time this month, I think I'd've been lynched - by both management and the users. :)
Sandy
If your users can recover to AREV, then issuing an EVAL UNLOCK ALL from the offending station should clear things up.