Timothy Murphy wrote:
I was doing a yum update (under FC-6) when the machine hung - while updating rekall-sql, to be precise - after a number of packages had been updated. I had to stop the process. I then ran "yum update" again, and it updated the remaining packages, and cleaned up after them. But the first lot of packages were never cleaned up. I wonder if there is any way of re-running the cleanup part of yum by itself, in these circumstances?
If you would have changed to the /var/cache/<repo>/packages directory after the incompleted yum transaction, then ram rpm -Uvh --replacepkgs --replacefiles *.rpm as root, the packages would have cleaned up the old database entries and assured the packages were installed and the scripts finished their tasks up. If you ran yum again after the fiasco, you probably need to download the packages that failed the transaction and perform as above.
CAUTION: make sure that there are no packages that install vs upgrade, like kernels in the directory before running the transaction with -Uvh
NOTE: bailing out of yum or rpm midstream usually leaves behind a mess. Jim -- Integrity has no need for rules.
