Re: How to debug anaconda add repo process?

[Chris Lumens <clumens redhat com>]

> 09:09:51 WARNING : ignoring duplicate repository updates with URL
> ['http://fedora.cora.nwra.com/fedora/linux/updates/10/x86_64/']

If your repo is named "updates", then it is a duplicate of the updates
repo mentioned in /etc/anaconda.repos.d/fedora-updates.repo.

> In the anaconda code, it just seems to
> assume that if a repo can't be added it's because it is duplicate:
> 
>         for repo in extraRepos:
>             try:
>                 self.repos.add(repo)
>                 log.info("added repository %s with URL %s" % (repo.name,
> repo.mirrorlist or repo.baseurl))
>             except:
>                 log.warning("ignoring duplicate repository %s with URL %s" %
> (repo.name, repo.mirrorlist or repo.baseurl))

When you see the add code in yum, it becomes obvious that there's only one
exception that can be raised:

    def add(self, repoobj):
        if self.repos.has_key(repoobj.id):
            raise Errors.DuplicateRepoError, 'Repository %s is listed more than once in the configur ation' % (repoobj.id)
        self.repos[repoobj.id] = repoobj


> So I don't have any information as to *why* we were not able to add repository.
>  Is there some way to enable verbose debug logging for the anaconda yum instance?

No, not really.

- Chris