On 07/19/2010 03:59 PM, David C. Rankin wrote:
Guys,
The 11.3 repo disaster has put me in a fix with my 11.3 install. Despite downloading the kernel 7 times, every time, it was corrupt. Along with mkinitrd, etc..
I have downloaded the rpms from the repo manually and I have used a usb stick to move them to the install box. However now, (when the install is stuck on another mismatched checksum), it will not let me 'rpm -Uvh new files' into the system. The error is:
error: cannot open Package index using db3 - no such file or directory (2) error: cannot open package database in /var/lib/rpm
I know this probably means I'm using the wrong rpm command, but I've never been in this position before.
Gurus -- what's the trick?
Well, There is a bit of a silver lining to the upgrade from hell. Since it was an 'upgrade', even with the corrupt kernel digest/checksums, when the upgrade completed, the system fell-back and booted the 2.6.31 (11.2) kernel and I was able to recover. The key for the 'upgrade' was not to panic and try to patch things during the install. I tried throwing the new kernel in during the upgrade, but I guess yast puts a lock on the rpm database throughout the install -- or -- the rpm database is just symlinked from some other location during install. Dunno. Once the installer was done installing packages, there wasn't any time to patch things between the end of the download and the reboot. I just copied the kernel rpms back to the box after the reboot and they installed just fine. Curiously the kernel-pae-devel and kernel-syms were already installed, but the rest was no where to be found. After the 7G upgrade, booting to the system and then choosing upgrade-all shows another 1.1G to go. I'll pull all those in and hope it fixes all of the packages that got skipped during the initial upgrade. (however -- I have a sneaking suspicion that any corrupt packages that were ignored instead of skipped will be on the system without a way for the installer to tell they need to be replaced (good thing I took notes on my notepad during the install :) The 11.3 install is impossible to compare to any of the other installs 8.0-present due to the redirector screw-up. With that mess, it took 19 hours to do a simple 11.2 -> 11.3 upgrade due to having to cancel all the corruption dialogs and then 'retry' to restart the download. (this is one time downloading of the DVD would have been smart along with a complete fresh install) It will take a few days -> week to get a feel if this install will make it through this operation or whether we should just call 'code' now and move on to the wipe->reinstall instead of upgrade. I'll see if I can collect the packages in my local repo so any reinstall is a 30m. orera My current update on the box is already hitting "wrong digest problems" so those issues are not limited to the install/upgrade setting. Good luck to the rest making the jump. Word to the wise -- wait a few days until the redirector issue is resolved and Marcus gives the all clear or you are in for a -- VERY VERY long update... I'll follow up with the results of the manual kernel install after the current update completes, but I don't expect any problems. The only anticipated issue is the radeon/2.6.34 bug that may/may not arise depending on how suse is handling the KMS early/late config. If it's KMS early, loading the radeon driver in the initramfs, then I'll be dead in the water and the kernel will crash when it gets to the "Loading Modules" line. If it is KMS late with the radeon module loaded later, boot should be fine, but I expect compiz to whitescreen. We shall see... -- David C. Rankin, J.D.,P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org