On Thursday 23 January 2020 15:22:32 Simon Matter via CentOS wrote:
Before you try the update again, you'll have to fix the reason for the failure - add memory, or at least add a swap file.
You could check with rpm -qa --last | head -20 which the latest packages are that were installed.
If the rpm database is corrupted, rebuild it with rpm --rebuilddb
You can reinstall packages that may not be completely installed, using rpm --reinstall PACKAGE_FILE
Well, there may be more to check and above steps may not help.
Do you have duplicates in rpmdb? Which one of the duplicates are already on the filesystem?
I do not believe that I have any duplicate RPM's installed, and I do not believe that the database is actually wrong. I believe that it is purely that the installs did not complete successfully.
I have managed to remove the latest kernel RPM and 'yum update kernel*' re-applied the update without errors.
I am going to go through the list of failed RPM's and to a 'rpm --reinstall' for each one in turn and hopefully they will work too.
Before doing that I will do as Kay suggested and find out why I had the memory problem in the first place.