Red Hat Bugzilla – Bug 76690
shortage of uploaded packages
Last modified: 2007-04-18 12:47:56 EDT
From Bugzilla Helper:
User-Agent: Mozilla/5.0 Galeon/1.2.6 (X11; Linux i686; U;) Gecko/20020913
Description of problem:
No matter how many times I create, populate, delete a new base
channel, I only get 25 packages identified as uploaded. The first time
there were close to 900, which is still many many hundred short.
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. upload packages
2. click 'clear cache' link on website when viewing channel contents
Actual Results: 25 packages appeared.
Expected Results: ~1700 packages should have appeared.
Joe, can you reproduce this problem? It looks like it might be a pagination
problem, but to test you'll need to load a whole bunch of custom packages. When
you can reproduce, hand off to Robin for troubleshooting.
Joe cannot replicate the problem. Isaiah, do you get error messages when you
sue the rhn_package_manager? Any diagnostic info that you can share?
I get no error messages. rhn_package_manager does mention the package has
already been submitted, after it uploads it. On the website, I click the
refresh content option on the page, and the 'click this when you're done
modifying the contents of the channel' link. I get 25 packages. The base
channel name is cisco-linux-i386-7.1 if you're interested.
What command string are you using to upload the packages? I used
rhn_package_manager -v --nosig -d /root/5.2 -c joe-base --username=jkomenda
Also, rpm -qa | grep rhn would probably be helpful.
Everything is up to date:
$ rpm -qa | grep rhn
The only applicable update to the 7.2 system is mod_python, which the
documentation made quiet clear should not be applied.
Here is my command:
sudo rhn_package_manager -v -c "cisco-linux-i386-7.1"
The contents of the directory are all 7.1+powertools packages.
Pulled directly from the 7.1 + powertools channels, right? Then these packages
are already in the database with the same NVRE and MD5sum, correct? I would
presume that the import of those packages would fail, and we should see some
errors *somewhere*. Todd, can you take a look at this?
I didn't pull them from the channel, they're from an FTP site mirror. But the
packages are still the same.
When run without -v, _no_ output is yielded.
New proxy updates (today) fixed this.