Bug 12796 - Possible bug in mgaapplet-update-checker
Summary: Possible bug in mgaapplet-update-checker
Alias: None
Product: Mageia
Classification: Unclassified
Component: Release (media or process) (show other bugs)
Version: 4
Hardware: x86_64 Linux
Priority: Normal minor
Target Milestone: ---
Assignee: Mageia Bug Squad
QA Contact:
URL: ftp://mirrors.kernel.org/mageia/distr...
Depends on:
Reported: 2014-02-17 03:19 CET by Robert Munro
Modified: 2015-04-03 12:26 CEST (History)
3 users (show)

See Also:
Source RPM: mgaonline-3.9-1.mga4.noarch.rpm
Status comment:


Description Robert Munro 2014-02-17 03:19:03 CET
Description of problem:
There might be an elusive bug in mgaapplet-update-checker, which is a component of mgaonline. When it checks the inactive mirror repositories, it fails with the message "Failed to open urpmi database" when it hits the Tainted 32bit Backports Testing repository. This occurs on a system that is a clean install of Mageia 4, using the mirrors at mirrors.kernel.org.

Version-Release number of selected component (if applicable):

How reproducible:
This error occurs consistently and automatically when mgaapplet-update-checker runs on the afflicted system. It shows on syslog and in a taskbar popup message.

See https://forums.mageia.org/en/viewtopic.php?f=7&t=7017 . There is attached file mirrorfix.txt at the end of that thread.

This might be as simple as mgaapplet-update-checker complaining that the named mirror repository Tainted 32bit Backports Testing is empty. When rsync copies that repository, it might not create the rpmdb file on the target if it's empty, however the synthesis.hdlist.cz file exists and appears on the affected system.


Steps to Reproduce:
Comment 1 Manuel Hiebel 2014-02-17 17:20:46 CET
Thierry, for you or sysadmin ?

CC: (none) => thierry.vignaud

Comment 2 Robert Munro 2014-03-03 00:16:57 CET
This glitch likely was triggered by the way I did the install. I used a hard disk iso image *and* mirror respositories, thinking that would be faster than doing a full network install. It was faster, but I ended up with both the hard disk Core Release and Nonfree repositories duplicated, so I had both Core Release and Core Release2, etc., in my configuration. I deleted everything and reselected just the mirror repositories, but then this problem appeared.

I have since done a network install and this problem did not reappear after that.
Comment 3 Rémi Verschelde 2015-04-03 12:26:12 CEST
Closing as invalid then (feel free to reopen if you consider there is still a usability issue that should be fixed though).

CC: (none) => remi
Resolution: (none) => INVALID

Note You need to log in before you can comment on or make changes to this bug.