One thing I can suggest (although it's not a natural process), is to always Log Off before shutting down the machine, this will flag up processes that have an 'end task' box so you can click on them before the vnc server stops you accessing the machine. This is fine for standard NT workstations as they have a Shutdown box, or servers with the 'any user can shutdown' policy, or you have to log in again to shutdown, adding that element of 'stuck process' risk again...
Of course if there's any services or other problems that interfere with the proper shutdown procedure then it's jumping time. Naturally it would be nice for someone to have a real good think about the way VNC hooks into the display driver (reason I've been told that it can't be done easily) and ensure that it cannot be shut down until almost the last thing before reboot. Cheers, Keith. > -----Original Message----- > From: Matt Swift [mailto:[EMAIL PROTECTED]] > Sent: Tuesday, March 05, 2002 10:46 PM > To: [EMAIL PROTECTED] > Subject: shutdown gotcha with Win32 host > > > When running VNC as a service on Win2k, one can reboot the host > machine via the remote client. The machine reboots, the service runs, > and one can reconnect with the client, log in, and so on -- > > *unless* during the shutdown process Windows encounters a program that > it cannot close (most notably, any Cygwin bash shell). Windows then > puts up a popup window asking the user to confirm a more forceful kill > of that program. The sequence of events in this case is the > following: > > select reboot on host > VNC host closes client connection > Windows encounters a problem process and requests user input > > In this state, the VNC host responds to attempts to reconnect with a > client, but after authentication immediately drops them, with a > message to the effect that the connection was dropped because a > shutdown is in process. > > The result is the inconvenient necessity, when rebooting a host > remotely, to kill as many processes as possible manually, to avoid the > possibility of locking oneself out of the system permanently, until > one can visit the actual remote machine and manually confirm to > Windows that it may kill that dratted Cygwin shell (or other process). > > There is more than one way to address this problem -- and probably > more than one way ought to be pursued. > > Is there a way to invoke an "unconditional reboot" in Win2k that I do > not know about? > > There will always be a risk of a "stuck" process, so if there is not a > way to ask Windows to reboot unconditionally, then I think VNC ought > to let clients connect when a shutdown is in process. I do not > understand the reason it declines to accept connections, but whatever > it is, I would be surprised if it applies in the circumstances > described, and would be surprised if the negative consequences of > allowing a connection are as dire as the consequences of declining > one. > > Any insight or suggestions will be greatly appreciated. > --------------------------------------------------------------------- > To unsubscribe, mail [EMAIL PROTECTED] with the line: > 'unsubscribe vnc-list' in the message BODY > See also: http://www.uk.research.att.com/vnc/intouch.html > --------------------------------------------------------------------- --------------------------------------------------------------------- To unsubscribe, mail [EMAIL PROTECTED] with the line: 'unsubscribe vnc-list' in the message BODY See also: http://www.uk.research.att.com/vnc/intouch.html ---------------------------------------------------------------------
