Bug 11542 - urpmi.update spends useless time on disabled medias
Summary: urpmi.update spends useless time on disabled medias
Status: RESOLVED OLD
Alias: None
Product: Mageia
Classification: Unclassified
Component: RPM Packages (show other bugs)
Version: Cauldron
Hardware: All Linux
Priority: Normal normal
Target Milestone: ---
Assignee: Thierry Vignaud
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2013-10-26 16:20 CEST by Thomas Backlund
Modified: 2015-05-17 00:53 CEST (History)
3 users (show)

See Also:
Source RPM: urpmi
CVE:
Status comment:

Attachments
Add an attachment (proposed patch, testcase, etc.)

Description Thomas Backlund 2013-10-26 16:20:51 CEST 
As my local server is currently down for rebuild I use online mirrors,
and noticed a very annoying behaviour...


urpmi.update "background checks" keeps spending time on disabled medias wich is bad:

[root@localhost ~]# LC_ALL=C urpmi --auto-update
urpmi database is locked, process 28170 is already using it (/usr/bin/perl /usr/libexec/urpmi.update Tainted Backports)

[root@localhost ~]# LC_ALL=C urpmq --list-media active
Core Release
Core Updates
Nonfree Release
Nonfree Updates


Now this get even more problematic on 64bit installs that has all 32bit medias too, so it takes _ages_ before urpmi  is usable again if the selected mirror happends to be slow responding.

It should only check/update active medias

Reproducible: 

Steps to Reproduce:
Comment 1 Manuel Hiebel 2013-10-26 18:48:18 CEST 
this could explain https://bugs.mageia.org/show_bug.cgi?id=11478
Comment 2 James Kerr 2013-10-26 22:55:16 CEST 
mgaapplet updates disabled backport repo's. IIRC,this is done so that the second filter in rpmdrake can display backports when the backports repo's are not enabled. In the Mandriva days there was a (lengthy) discussion about how backports should be handled by rpmdrake:

https://qa.mandriva.com/show_bug.cgi?id=40556
Comment 3 James Kerr 2013-10-26 23:07:03 CEST 
Correction - it's the first filter in rpmdrake that is intended to display backports.
Comment 4 Thomas Backlund 2013-10-26 23:18:48 CEST 
Yeah, but from what I saw it seems to probe every disabled mediam including debug, testing, 32bit, ...
Comment 5 Dave Hodgins 2013-10-26 23:25:12 CEST 
Workaround would be to remove all unwanted media, instead of just disabling them.
urpmi.removemedia -y BackPorts
Same for Debug, Testing, and for most cauldron users Updates.

CC: (none) => davidwhodgins

claire robinson 2013-10-28 08:43:30 CET

CC: (none) => eeeemail

Comment 6 Marja Van Waes 2013-11-22 15:16:53 CET 
Is this bug still valid?

Until today I had to kill 2 instances of mgaapplet and then urpmi.update (first killing urpmi.update didn't work, because it was started again right away), before I could run "urpmi --auto-update". 

The problem seems solved now. My uptime is 32 minutes, only one mgaaplet is running (sleeping now) and no urpmi.update

I don't get the message about the locked urpmi database any more

CC: (none) => marja11

Comment 7 Samuel Verschelde 2015-05-17 00:53:17 CEST 
No answer to comment #6, closing.

Status: NEW => RESOLVED
Resolution: (none) => OLD
Note You need to log in before you can comment on or make changes to this bug.