Hello, Some time ago, I upgraded my system from RedHat 5.1 to 6.0, and I experienced some troubles because of my rather "unusual" partition settings. Since I just read you are working on a new release, I describe you the problem and hope it will be taken care of to make the new RedHat upgrade even a bit better. The problem was that the update script didn't find the RPM database located somewhere in /var. It didn't find it because /var was just a symlink to another directory on a seperate partition, which isn't mounted by the upgrade script. It took some time before I understood what was going on, afterwards I just moved the real /var directory to my root partition and moved it back to the seperate partition after the upgrade process. I suggest you would at least document this situation, or better, create an option to mount the needed partition during the upgrade. A last note: I believe the upgrade script complained about the /tmp directory as well - also a symlink - but this problem wasn't fatal for the upgrade itself. Kind greetings, Tung <tg>
thanks for your report, tung! Have you experienced the same problems with 6.2?
The current internal release (Winston, beta2) more gracefully brings up an RPM db dialog error when this situation occurs: 'Rebuild of RPM database failed' We will add your request for even better handling of this situation to our feature request list ... thank you for your report!
Related to bug 18092.
*** This bug has been marked as a duplicate of 13071 ***