Bug 520179
Summary: | yum downloads files from file:/// repositories before installing them | ||
---|---|---|---|
Product: | [Fedora] Fedora | Reporter: | Chuck Ebbert <cebbert> |
Component: | yum | Assignee: | Seth Vidal <skvidal> |
Status: | CLOSED WORKSFORME | QA Contact: | Fedora Extras Quality Assurance <extras-qa> |
Severity: | medium | Docs Contact: | |
Priority: | low | ||
Version: | rawhide | CC: | ffesti, james.antill, maxamillion, pmatilai, tim.lauridsen |
Target Milestone: | --- | ||
Target Release: | --- | ||
Hardware: | All | ||
OS: | Linux | ||
Whiteboard: | |||
Fixed In Version: | Doc Type: | Bug Fix | |
Doc Text: | Story Points: | --- | |
Clone Of: | Environment: | ||
Last Closed: | 2009-10-12 21:21:14 UTC | Type: | --- |
Regression: | --- | Mount Type: | --- |
Documentation: | --- | CRM: | |
Verified Versions: | Category: | --- | |
oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
Cloudforms Team: | --- | Target Upstream Version: | |
Embargoed: |
Description
Chuck Ebbert
2009-08-28 18:50:37 UTC
Also happens on F-11 with yum-3.2.23-3.fc11 and python-urlgrabber-3.0.0-15.fc11 I cannot replicate this on any version of yum right now. Could you post all of your .repo files and your yum.conf just to be sure: do you have an /etc/yum.repos.d dir AND an /etc/yum/repos.d dir? thanks (In reply to comment #2) > I cannot replicate this on any version of yum right now. Could you post all of > your .repo files and your yum.conf just to be sure: > > do you have an /etc/yum.repos.d dir AND an /etc/yum/repos.d dir? no. I straced yum and it stats the .rpm packages it's going to install. Then it opens /proc/mounts and reads it, then stats all the mounted ext3 and ext4 filesystems. It really looks like it's trying to see if the file:/// URL corresponds to a mounted local filesystem, but seems not to understand that cifs is one of them. just for the sake of argument - can you setup a truly local repo and test it so we can be certain this is a cifs thing and not something else? (In reply to comment #4) > just for the sake of argument - can you setup a truly local repo and test it so > we can be certain this is a cifs thing and not something else? I tried it from a locally mounted ext2 filesystem and it still downloaded the packages before installing. I did notice two things though, that are the same on the CIFS and local mounts: 1. The repo is owned by a user other than root. 2. There is a drpms directory but it's empty because I don't mirror that. The owner of the rpms makes no difference whatsoever. Please disable the presto plugin and re-run your test. Chuck - a reping on the thing from last week - please disable the presto plugin and re-run your test. Thanks, I can't replicate this - and neither can anyone else afaict. if you can do the tests from above that would help reopen then yum-presto is not even installed on one of the machines where this happens Are any of these machines rhts machines? Is there any way I can take a look b/c I have no idea what's happening here. okay - I've setup a cifs share using samba3, mounted it using mount and added it as a repo with the repo config of: [cifs-empty] name = cifs baseurl=file:///mnt/cifs-test/tmp/empty-repo enabled=0 gpgcheck=0 I enabled the repo and I installed one pkg from this repository. The package was accessed directly from this location, it was not downloaded to /var/cache/yum/.... So I cannot make this one happen at all. Any suggestions on how to duplicate this bug? |