On Sat, 2005-07-23 at 00:29 -0400, Michael Wiktowy wrote: > > Date: Fri, 22 Jul 2005 12:28:27 -0400 > > From: Bob Chiodini <rchiodin@xxxxxxxxxxxxx> > > Message-ID: <1122049707.7272.11.camel@xxxxxxxxxxxxxxxxxxx> > > Content-Type: text/plain > > Mike, > > > > If it happened here a little more often I'd try sniffing the network to > > see if some kind of protocol error was occurring. I believe that > > connector accesses Exchange via it's webmail interface. I suspect that > > some http error is occurring that crashes the data server. Usually > > closing and rerunning evolution gets good results. > > > > I've always believed that the data server was spawned when evolution > > started, if it was not running already. I do not think I have ever seen > > the data server running after a reboot, but before running evolution. > > > > As a test, I just killed the data server with signal 11 (SIGSEG) and was > > presented with the "inform developers" dialog, but no information was > > available in the debug info option, even after selecting update. Try > > running the "inform developers" option and see if you can get it to > > forward a bug report to Novell. Maybe installing the evolution-data- > > server-debuginfo rpm will help > > (http://mirror.linux.duke.edu/pub/fedora/linux/core/updates/3/i386/debug/ for FC3). > > > > As another point, when I was running connector under evo 1.4, rapidly > > switching between email and the calendar crashed the data-server every > > time. > > > > Bob... > > Bob, > I was getting the Inform developers dialog quite regularly. > I did figure out what was wrong though. Apparently, e-d-s freaks out > completely when it can't find its calendar files. I had subscribed to > some remote calendars and then the remote server changed names. That > caused Evo to lock up solid when I switched to calendar view. It stuck > in calendar view from then on (and consequently locking up) until I > opened evo from the command line and used the mail flag. > I had to go into gconf-editor and fix the broken remote calendar > references. Now everything is smooth again. I will definitely bugzilla > this upstream. I suspect this might be your source of the odd crash > also. Perhaps your Exchange server is returning the odd timeout to Evo > causing it to die. > > ref: http://bugzilla.gnome.org/show_bug.cgi?id=311313 > > Not sure if this is corrected in FC4/Gnome 2.10. > > /Mike > Mike, Thanks for the info. Agreed on the odd timeout. Bob...
