I've been away from RedHat-based distros for a while, but I need to test a commercial app which requires FC3 or FC4, so today I set up an FC4 machine using .iso's freshly downloaded from a mirror. I did a base "server" install and all seemed to go well, but somehow I managed to leave out X11 support, which our app needs, so I'm trying to install it using up2date. Well, just about anything I try to do with up2date fails, and each time I try I tend to get different errors. A few samples: " There was some sort of I/O error: <urlopen error (110, 'Connection timed out')> " " There was some sort of I/O error: <urlopen error (104, 'Connection reset by peer')> " " There was a fatal error communicating with the server. The message was: An HTTP error occurred: URL: http://linux.admin.uillinois.edu/pub/linux/fedora/linux/core/updates/4/i386///openssl-0.9.7f-7.10.i686.rpm Status Code: 403 Error Message: Forbidden " ...and so on. This happens whether I try to do "up2date --nox --list", "up2date --nox --install xorg-x11", "up2date --nox -u", or just about anything else. I can't find any pattern to the errors; trying the same thing twice in a row has yet to produce the same error twice. Sometimes the error message directs me to /var/log/up2date, for example: [...] initRepo bytes=440-35945 initRepo bytes=440-25242 An error has occurred: exceptions.IndexError See /var/log/up2date for more information But in that log file I find what looks to me like a Python traceback, but nothing in it suggests anything meaningful to me. Since I've gotten lots of 403 and 404 errors, I wonder: is there a problem currently with a lot of Fedora mirrors? I've gotten these from several different servers, so I don't think it's one corrupt mirror someplace. When I do get errors from the same server twice, it tends to be complaining about two different URL's. I thought I might be tripping over mirrors in the midst of re-seeding themselves or something, but the trouble has been persisting now for several hours. I've tried reinstalling the system from scratch, thinking there might have been some database corruption or some such; no difference. So now I'm stuck. Anybody have any idea what might be wrong? Thanks, --michael
