[Yum] Poll: How users use DNF

Bruno Wolff III bruno at wolff.to
Tue Dec 9 17:48:00 UTC 2014


On Tue, Dec 09, 2014 at 12:23:56 -0500,
  Radek Holy <rholy at redhat.com> wrote:
>Dear users of YUM and DNF,
>
>Please share with me the use cases, not the description of the "install" command. Think twice before you share something because I believe it's not as easy as it might seem. As an example I think it might be something like:

I use it to keep up very up to date with local mirrors and to put back 
packages that I have removed to work around conflicts (typically from 
soname bumps).

I keep a list of packages I have removed in a file.

After I refresh my local mirror I rumn a script that does the following:

clean all so that the meta data will be refreshed.

An update to get the latest packages.

Try to install all of the packages that had previously been removed.

Then another update so I can more easily see which updates are failing 
(typically do to soname bumps).

Then I use package-cleanup to get a list of orphans.

After I run the base script, I'll usually look to clean up orphans and 
deal with packages that are not updating. yum update -y -v | grep -i fail 
is a good way (but not 100% reliable) to find packages blocking library 
updates.

After that clean up is done if it looks like some packages on my removed 
list have been reinstalled (or deleted from a repo), I'll run 
yum list avaiable `cat putback` to get an up to date list of packages 
I need to reinstall.

I use list and search to find packages.

I modify my repo.conf file to include optional packages in groups, enable 
skip broken and disable delta rpms (with a full local mirror delta rpms 
don't help).

I use distro-sync to downgrade packages that disappear from updates-testing 
or occasional downgrades that shouldn't happen in rawhide.

I use yum to upgrade between fedora versions. This can be a bit of a mess 
that usually works best if I find out which packages have been dropped 
that I have installed before doing the upgrade, so that I can remove them 
and then try to install them later to pick up obsoletes. Also I usually 
temporarily remove packages that have conflicts in the new release if 
I know about them in advance.

I use repoquery to look for stuff from time to time. Those uses are pretty 
ad hoc and I really don't have anything standard I do.

I occasionally use yum in a chroot while doing a rescue after a bad update 
has made a system unbootable. I have to make sure that I bind mount the 
special directories, but this works pretty well.

When I add new packages that I want I will typically use yum to do the 
install rather than the GUI way to do this.


More information about the Yum mailing list