On 11/23/06, Gérard Henry <[EMAIL PROTECTED]> wrote:
yes, that's right.I have 3 users that use JDS (others use CDE), and only one user (same user!) encounter many troubles.
The most likely cause is one of the things Bob Doolittle mentioned early in this thread: an xscreensaver "hack" (one of the programs that runs in order to waste CPU cycles and network bandwidth by drawing to the screen while xscreensaver has the session locked) could be doing something that triggers a bug in the X server and causes it to crash. If the X server goes away, the desktop goes with it. If this user has xscreensaver set to choose hacks at random then that might explain why sometimes his session lives and sometimes it dies. There are three thngs you can do: - after this user's session has disappeared, look in /var/dt/Xerrors for messages that indicate that his X server has died unexpectedly. Sun Ray X servers write a register dump into Xerrors in some cases, but dtlogin should always write some sort of message if an X server dies before its session logs out. - configure the system to capture core dumps from X servers that die violently. X servers are setgid programs so they don't usually leave core files behind, but you can use 'coreadm' to tell the system to capture those core files into a specific directory. - or you can try to just avoid the problem by making sure that this user configures his xscreensaver to run only the simplest hacks. Even better, configure it to run no hacks at all and just blank the screen while the lock is active. One easy way to do that is to 'pkgrm' the xscreensaver-hacks package. (The Sun X team deliberately placed the hacks in a separate package so that they can easily be removed from the system.) It's still possible for people to get around this by installing their own hacks but very few people will bother to do that. I think you can prevent even that from happening by creating "mandatory" gconf settings for xscreensaver but I don't have instructions handy for that. They might be on the ThinkThin blog. Doing the first two, at least for a while, would give us the best chance of finding and fixing the cause of the problem. OttoM. __ ottomeister Dsclaimer: These are my opinions. I do not speak for my employer. _______________________________________________ SunRay-Users mailing list [email protected] http://www.filibeto.org/mailman/listinfo/sunray-users
