On Saturday 04 February 2006 23:04, Jim Cornette wrote:
>Gene Heskett wrote:
>> On Saturday 04 February 2006 20:50, Jim Cornette wrote:
>>> Danny Terweij - Net Tuning | Net wrote:
>>>> Hi there,
>>>>
>>>> While running a yum update the hdd did get an error and the system
>>>> was remounted in read only mode.
>>>>
>>>> The error seems gone, but the rpmdb is not fine anymore, i did do
>>>> rpm --rebuilddb
>>>> But now i get weird errors with installing packages, like dupes or
>>>> already installed (but it is not).
>>>>
>>>>
>>>> How do i create a new rpmdb with current installed packages?
>>>>
>>>> Or am i doomed? :)
>>>>
>>>> Danny
>>>
>>> Basically, you have to find the multiple version installed rpms and
>>> use a option to rpm called --justdb. this option will remove only
>>> the database entry for the old nonexistent package entry and then
>>> you could use the -V or --verify option to rpm to ensure the later
>>> installed version is intact. No output should show on the verify.
>>>
>>> Someone on the test list posted this one-line command to allow an
>>> output to the terminal which will detect multiple entries of
>>> packages. The kernel and the gpg-pubkey and any other package that
>>> allows for multiple versions to coexist. If you are running a
>>> 64-bit system, the command will not work well.
>>>
>>> rpm -aq --queryformat "%{NAME}\n" | sort | uniq -c | grep -v -E "
>>> *1 "
>>>
>>> For packages that are not actually installed but are in the rpm
>>> database, it might be a tougher issue to deal with. You would need
>>> to run something like 'rpm -qa |grep missing' as root to detect
>>> packages which have missing files. Then you can either download the
>>> rpms and use rpm -Uvh --replacefiles --replacepkgs <package.rpm>
>>> or remove the database entry for the messed up package and do a
>>> 'yum install <package>' on the shortchanged package. Yum should not
>>> know it is installed and install over whatever remained of the
>>> package.
>>>
>>> This other suggestions or you're doomed as you stated above. :-)
>>>
>>> Jim
>>
>> Hmm, jumping in here, that command above returns this:
>> [root@coyote example]# rpm -aq --queryformat "%{NAME}\n" | sort |
>> uniq -c | grep -v -E " *1 "
>> 2 gnome-mag
>> 2 gnome-speech
>> 28 gpg-pubkey
>> 2 kernel
>> 2 perl-Digest-HMAC
>> 2 perl-Digest-SHA1
>> 2 perl-Time-HiRes
>>
>> Frankly, with as much stuff as I've built and installed from
>> tarballs, using either the make install or the checkinstall option,
>> I would have thought the output would have been much more verbose. I
>> use kde, self installed, so those gnome entries could probably be
>> removed. Is there a way to get the versions back from that command
>> automaticly?
>
>There was a script out there that someone made that made a file in the
>/tmp directory, then compared the uniq version and outputted the
> version information for the multiple versions. I long lost the script
> but it is in the archives on the fedora-test-list.
>I guess you could add rpm -q on the short list with the packagename
>files. The exact versions should show using the rpm -q function.
>
>I should learn scripting instead of using ready-made scripts. Some day
>maybe.
>
>Jim
;-) I hear that, Jim. Me, at 71 I tend to furget, so I write scripts
that don't forget. Yup, been there, done that. Several times... And a
month later I may have forgotten the scripts name, and where I stashed
it. I've heard that called CRS.
--
Cheers, Gene
People having trouble with vz bouncing email to me should add the word
'online' between the 'verizon', and the dot which bypasses vz's
stupid bounce rules. I do use spamassassin too. :-)
Yahoo.com and AOL/TW attorneys please note, additions to the above
message by Gene Heskett are:
Copyright 2006 by Maurice Eugene Heskett, all rights reserved.
