Bug 17937 - Cannot add online media after install from Classical DVD "unable to open rpmdb"
Summary: Cannot add online media after install from Classical DVD "unable to open rpmdb"
Status: RESOLVED WORKSFORME
Alias: None
Product: Mageia
Classification: Unclassified
Component: RPM Packages (show other bugs)
Version: Cauldron
Hardware: x86_64 Linux
Priority: Normal normal
Target Milestone: ---
Assignee: Mageia Bug Squad
QA Contact:
URL:
Whiteboard: 6dev1
Keywords:
Depends on:
Blocks:
 
Reported: 2016-03-10 00:36 CET by Thomas Andrews
Modified: 2016-03-13 03:42 CET (History)
4 users (show)

See Also:
Source RPM:
CVE:
Status comment:


Attachments

Description Thomas Andrews 2016-03-10 00:36:47 CET
Description of problem:
When trying to add online media after Plasma 5 install, MCC fails with an "unable to open rpmdb" error message. Also, attempting to access software on the iso fails with the same message.

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

March 6 build of the Classical x86-64 DVD, in VirtualBox.
Problem may exist in a 32-bit install as well, but I was unable to get that far.

How reproducible:


Steps to Reproduce:
1.
2.
3.
Thomas Andrews 2016-03-10 00:37:27 CET

Whiteboard: (none) => 6dev1

Comment 1 Marja Van Waes 2016-03-10 10:50:49 CET
That means the rpm database got corrupted.

Running (as root) the following commands in the below order, should fix it.

 rm -f /var/lib/rpm/__db*
 rpm --rebuilddb



@ Thierry

would it make any sense, when this error occurs, to let urpmi auto-invoke removal and rebuild of those databases?

CC: (none) => marja11, thierry.vignaud

Comment 2 Marja Van Waes 2016-03-10 12:22:52 CET
s/urpmi/whichever program gave the error/
Comment 3 Thomas Andrews 2016-03-10 17:53:58 CET
(In reply to Marja van Waes from comment #1)
> That means the rpm database got corrupted.
> 
> Running (as root) the following commands in the below order, should fix it.
> 
>  rm -f /var/lib/rpm/__db*
>  rpm --rebuilddb
> 
> 
> 
> @ Thierry
> 
> would it make any sense, when this error occurs, to let urpmi auto-invoke
> removal and rebuild of those databases?

Yes, that seems to have done it. Adding a specific mirror as I type...
Comment 4 Marja Van Waes 2016-03-12 20:13:06 CET
(In reply to Thomas Andrews from comment #3)

> 
> Yes, that seems to have done it. Adding a specific mirror as I type...

Closing then, I forgot what all can cause database corruption, but do know there is nothing we can do to prevent it. Glad the workaround helped :-)

@ LpSolit @ Stormi

In cases like this, I'd like to have another resolved option.

Tbh, I prefer "CANTFIX" over "WONTFIX"

WDYT?

Status: NEW => RESOLVED
CC: (none) => LpSolit, stormi
Resolution: (none) => FIXED

Comment 5 Frédéric "LpSolit" Buclin 2016-03-12 20:38:49 CET
(In reply to Marja van Waes from comment #4)
> In cases like this, I'd like to have another resolved option.
> 
> Tbh, I prefer "CANTFIX" over "WONTFIX"

WORKSFORME (works for me) would be better than FIXED in cases like this one, IMO. I'm not sure another resolution would help. But this is not the right place for this discussion. :)
Comment 6 Marja Van Waes 2016-03-12 20:51:58 CET
(In reply to Frédéric Buclin from comment #5)

> WORKSFORME (works for me) would be better than FIXED in cases like this one,
> IMO.

You're right, don't know why I (nearly) never chose that solution. Thx :-)

Resolution: FIXED => WORKSFORME

Comment 7 Thomas Andrews 2016-03-13 03:42:39 CET
I agree, too. "Fixed" in this case didn't really apply. 

I'll hope this was a one-time thing, but I'll reserve the right to open it up again should the symptom persist in future isos.

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