Bug 22081

(In reply to Simon Geard from comment #0)

> So now I can't install anything at all. Some of the failures a couple of
> months ago were due to the / partition running out of space so maybe one (or
> more) of the /var/rpm files has become corrupt but I don't know how to find
> that out. There is now plenty of space (11G).

There is a global option in
mcc->Software Management->Configure media sources for install and update-software

to Verify RPMs to be installed:
While there I would suggest setting downloader to wget.


> Could it be waiting for an internet location that no longer exists, if so
> how do I identify that location?

I can not answer that question as asked. But, http://mirrors.mageia.org/ has mirror location and bandwidth columns. Pick a few you might be interested in and go to http://mirrors.mageia.org/status to check their status.

Then you can get back into Configure media sources for install remove current selections, File: Add a specific media mirror to set the mirror you have decided to use.

As for downloading updates, I wrote/use a bash script to pull down all rpms to be installed and have it run a install test check. If OK, run it again without the --test switch.  

I find it handy to use the "script" command to watch what is going on and have a log which can be checked to see if there were problems with download or other snags. Commands would be:

script -c "urpmi --downloader wget --replacefiles --auto-update --auto --download-all --wait-lock --test"  pull_updates.log

If no problems in pull_updates.log, run


script -c "urpmi --downloader wget --replacefiles --auto-update --auto --download-all --wait-lock"  install_updates.log

install_updates.log should allow you to zero in on any rpm problem install.

If you write a script, I can recommend running something like
script --return --append  --quiet --command \
           "urpmi.update --wget $_wget_options -a --wait-lock" urpmi.log
which would sync your synthesis.hdlist.cz files with the mirror prior to pulling down your updates.

There is no sense in doing updates if the urpmi.update command has problems.

CC: (none) => bittwister2

Thank you for your prompt response. I'll investigate as you suggest and report back.

(In reply to Simon Geard from comment #0)

> using mirror
> http://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64
> examining synthesis file [/var/lib/urpmi/Core Release2/synthesis.hdlist.cz]

<examining more Mageia systhesis files>

> examining synthesis file [/var/lib/urpmi/google-chrome/synthesis.hdlist.cz]

<snip>

> Could it be waiting for an internet location that no longer exists, if so
> how do I identify that location?
> 

It seems to hang on the google-chrome synthesis file. The google-chrome medium isn't in a Mageia mirror, but I only see in your output that a Mageia mirror is used, there is nothing about using a Google mirror.

If you followed Bit Twister's instructions, then you have disabled or (better) removed all current media in this screen http://doc.mageia.org/mcc/6/en/content/drakrpm-edit-media.html and then added a new mirror.

In that process, the google-chrome medium should have been disabled or removed, too.

Keywords: (none) => NEEDINFO
CC: (none) => marja11

(In reply to Simon Geard from comment #0)
> When I try to install or update software using drakrpm it hangs whilst
> trying to build the list of available packages. 

Do keep in mind, if there is a kernel package to be installed, grub2 os prober can take awhile mounting all your partitions, scan every kernel image found to generate a grub2 menu entry.

Does not hurt to have another terminal running
     journalctl --no-hostname -fa --no-pager
to keep an eye on what is going on, if anything, with the system.

If the system is not up and running 24 hours a day, 15 to 45 minutes after you have logged in, daily/weekly cron jobs will start running which may impact your install performance.

Thanks for the heads-up on journalctl.

I have tried to update my media but the system hangs on the 'fetching list' step so it feels as if I've made negative progress!

The journalctl is reporting something interesting, lots of messages of the form

Nov 26 11:28:03 mandi[2938]: received event for wlp3s0: <3>CTRL-EVENT-BSS-ADDED 6733 d2:05:c2:4f:51:31
Nov 26 11:28:03 mandi[2938]: received event for wlp3s0: <3>WPS-AP-AVAILABLE
Nov 26 11:28:03 mandi[2938]: received event for wlp3s0: <3>CTRL-EVENT-NETWORK-NOT-FOUND
Nov 26 11:28:03 mandi[2938]: received event for wlp3s0: <3>CTRL-EVENT-BSS-ADDED 6734 00:8e:f2:6a:1e:a0
Nov 26 11:28:03 mandi[2938]: received event for wlp3s0: <3>WPS-AP-AVAILABLE

which as I understand it relates to the wireless connection, and yet I have a wired connection with the wireless connection turned off.

Typically I run my system for hours a day and do not have many cron jobs setup.
Console output tracing from editmedia:

[root@localhost rpm]# /usr/bin/perl -Dt /usr/libexec/drakrpm-editmedia --no-splash
Recompile perl with -DDEBUGGING to use -D switch (did you mean -d ?)
Ignore the following Glib::Object::Introspection & Gtk3 warnings
Subroutine Gtk3::main redefined at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 525.
getting exclusive lock on urpmi
not using metalink since requested downloader does not handle it
using mirror http://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64
retrieving media.cfg file...
retrieved http://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64 media/media_info/media.cfg

(In reply to Simon Geard from comment #5)
> 
> which as I understand it relates to the wireless connection, and yet I have
> a wired connection with the wireless connection turned off.

If messages arrive while attempting the update, then wireless is not disabled.

> Typically I run my system for hours a day and do not have many cron jobs
> setup.

I understand, but there are jobs installed by default, and you making changes to msec security/audit settings can interact with whatever you are doing interactively. Try "locate /etc/cron." for a list of jobs on your install.

> Console output tracing from editmedia:
 <snip>
> http://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64
> media/media_info/media.cfg

At a glance you do not seem to be following instructions. The above seems to indicate you are using $MIRRORLIST instead of a static media mirror as I suggested.

Normally Bugzilla bugs are for reporting bugs, not for troubleshooting.

I can recommend, opening a thread at https://forums.mageia.org/en/ to get some help on troubleshooting your problem. Another help method would be getting a free Usenet account at http://www.eternal-september.org/ and subscribe to the alt.os.linux.mageia news group.

(In reply to Bit Twister from comment #6)
> (In reply to Simon Geard from comment #5)
> > 

> 
> > Console output tracing from editmedia:
>  <snip>
> > http://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64
> > media/media_info/media.cfg
> 
> At a glance you do not seem to be following instructions. The above seems to
> indicate you are using $MIRRORLIST instead of a static media mirror as I
> suggested.

See http://mirrors.mageia.org/mirrors/www.mirrorservice.org 
Mirrorservice.org is not owned by Mageia, it is located in the UK and is _not_ our mirrorlist ;-)
> 
> Normally Bugzilla bugs are for reporting bugs, not for troubleshooting.
> 
> I can recommend, opening a thread at https://forums.mageia.org/en/ to get
> some help on troubleshooting your problem. Another help method would be
> getting a free Usenet account at http://www.eternal-september.org/ and
> subscribe to the alt.os.linux.mageia news group.

Bugzilla is indeed not the best place to figure out whether a problem is a bug.

It would help if you'd do as Bit Twister suggests, Simon.

The reason that I didn't assign this bug report to a team to fix it, is that I'm not sure, yet, that there is a bug that causes the problem you have.

Apart from that, if it becomes likely there really is a bug, we'll need more information (to be _attached_ to this report), like a copy of:

  /etc/urpmi/urpmi.cfg

and we'd need you to reboot, let drakrpm fail again and then run, as root:

  journalctl -ab > log.txt

after which log.txt should be attached, too.

Feel free to already attach those files to this report, though, and to refer to this bug report when you ask for help in the forums or in the news group.

Setting this report to unconfirmed for now. Please do inform us of the result of the troubleshooting.

Thanks!

Status: NEW => UNCONFIRMED
Ever confirmed: 1 => 0

I've done some more work on this. I've explicitly set the URL in the urpmi.cfg (attached) and it does not download some files to install.
I have used

script -c "urpmi --downloader wget --replacefiles --auto-update --auto --download-all --wait-lock --verbose --test"  pull_updates.log

as suggested but wget is failing and even with --verbose added (log file attached):

error: wget failed: exited with 8

...retrieving failed: wget failed: exited with 8

error: wget failed: exited with 8

error: wget failed: exited with 8

...retrieving failed: wget failed: exited with 8

error: wget failed: exited with 8

Created attachment 9810 [details]
config file for urpmi

Created attachment 9811 [details]
log file from urpmi

(In reply to Simon Geard from comment #8)
> I've done some more work on this. I've explicitly set the URL in the
> urpmi.cfg (attached) and it does not download some files to install.
> I have used
> 
> script -c "urpmi --downloader wget --replacefiles --auto-update --auto
> --download-all --wait-lock --verbose --test"  pull_updates.log
> 
> as suggested but wget is failing and even with --verbose added (log file
> attached):
> 
> error: wget failed: exited with 8

$ man wget | grep ' 8 '
       8   Server issued an error response.

Server/internet problems are why I suggested running urpmi.update first.

script --return --quiet --command \
           "urpmi.update --wget $_wget_options -a --wait-lock" urpmi.log
if [ $? -ne 0 ] ; then
  echo "urpmi.update failed, check urpmi.log"
  exit 1
fi

 _count=$(grep --count -i --extended-regexp "error|warn|fail" urpmi.log)
if [ $_count -gt 0 ] ; then
  echo "check for error/failure/warnings in urpmi.log"
  exit 1
fi

You will have a much better chance of help if you were to follow suggestion(s) given in last paragraph in comment #6

(In reply to Simon Geard from comment #8)
> I've done some more work on this. I've explicitly set the URL in the
> urpmi.cfg (attached) and it does not download some files to install.
> I have used
> 
> script -c "urpmi --downloader wget --replacefiles --auto-update --auto
> --download-all --wait-lock --verbose --test"  pull_updates.log
> 
> as suggested but wget is failing and even with --verbose added (log file
> attached):
> 

The funny thing is, that it would probably have installed 12 packages if you hadn't run the command with "--test" ;-)


The mirror you selected, distrib-coffee.ipsl.jussieu.fr has very little bandwidth and is often overloaded, besides: it currently has an issue with the http protocol you are using (http is in red on the distrib-coffee line here http://mirrors.mageia.org/ and the mirror doesn't even show up in http://mirrors.mageia.org/status )

Please try with another mirror that has much more bandwidth (see the first page), and which is up-to-date (green in the last page).

I check both those sites yesterday and I've just checked them again before this posting and neither indicate any problems. However I have now changed to

ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia

which is in the UK and has bigger bandwidth.

I have now removed the --test option and installed the updates okay but I still don't understand if the error 8 reported by wget means something is still wrong.

I think I understand now. After enabling the non-free packages wget fails because it is using the wrong url, e.g.

...retrieving failed: wget failed: exited with 8

Installation failed, some files are missing:
    ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64/nvidia-current-devel-384.98-1.mga6.nonfree.x86_64.rpm

I thought that the with-dir entry in the .cfg file would be composed with the url to get the location but it isn't, to get it to work the complete url including the location within the repository must be specified. I tried to find information on the with-dir option but failed.

I have run urpmi.update:

[root@localhost urpmi]# urpmi.update -a
...retrieving failed: wget failed: exited with 8

...retrieving failed: wget failed: exited with 8

medium "Nonfree Release2" is up-to-date
medium "Nonfree Updates" is up-to-date

which is as I expected since I haven't yet updated the first two entries to include the location:

Core\ Release2 ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64 {
  key-ids: 80420f66
  mirrorlist: 
  with-dir: media/core/release
}

Core\ Updates ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64 {
  key-ids: 80420f66
  mirrorlist: 
  update
  with-dir: media/core/updates
}

Nonfree\ Release2 ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64/media/nonfree/release {
  key-ids: 80420f66
  mirrorlist: 
  with-dir: media/nonfree/release
}

Nonfree\ Updates ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64/media/nonfree/updates {
  key-ids: 80420f66
  mirrorlist: 
  update
  with-dir: media/nonfree/updates
}

(In reply to Simon Geard from comment #13)

> I have now removed the --test option and installed the updates okay

Yuck, depending on which was failing/passing I have managed to screw up my install in the past more than once. I will not proceed with an install unless there are no wget errors.

>  but I
> still don't understand if the error 8 reported by wget means something is
> still wrong.

It does. 


> I think I understand now. After enabling the non-free packages wget fails
> because it is using the wrong url, e.g.
> 
> ...retrieving failed: wget failed: exited with 8
> 
> Installation failed, some files are missing:

Yep, have seen that problem disappear when my wget errors disappeared.

 
> Core\ Release2
> ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64 {
>   key-ids: 80420f66
>   mirrorlist: 
>   with-dir: media/core/release
> }

FRAP, and yet again, failure to follow instructions is helping you into the ditch and increase the likely hood of possible problems/failure. :-(

My instructions to delete all the mirror selections was to explicitly get rid of any/all mirrorlist entries, and get a known valid mirror to be used as the source of the package database/rpms to be downloaded.

If done properly a "grep -i mirrorlist  /etc/urpmi/urpmi.cfg" would return nothing found, which was/is the desired goal.

mirrorlist logic will pick a site for you and you may have noticed http://mirrors.mageia.org/status shows not all sites are current.
I also suggest to you, you can/may have less errors picking http over ftp.

FYI: picking the fastest site will not get you a faster download unless you are paying for the same speed from your ISP.

Until you can get a clean run from the urpmi.update step, you are just wasting time and effort trying to get a clean update.

As far as I can see an empty mirrorlist is the same as no mirrorlist but I have removed all instances of it now from the cfg file and the result of an updated is the same:

[root@localhost urpmi]# urpmi.update -a
...retrieving failed: wget failed: exited with 8

...retrieving failed: wget failed: exited with 8

medium "Nonfree Release2" is up-to-date
medium "Nonfree Updates" is up-to-date

... and just to check for mirrorlist
[root@localhost urpmi]# fgrep -i mirrorlist  /etc/urpmi/urpmi.cfg
[root@localhost urpmi]# echo $?
1

So now I have no mirrorlist entries and am using a known valid mirror as per your suggestions or have I missed something?

It still looks as if the with-dir entry is not being honoured 

Core\ Release2 ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64 {
  key-ids: 80420f66
  with-dir: media/core/release
}

Nonfree\ Release2 ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64/media/nonfree/release {
  key-ids: 80420f66
  with-dir: media/nonfree/release
}

I would like you to use an editor and delete all the stuff after
the cdrom stanzas, if any, in /etc/urpmi/urpmi.cfg

Personally, I would delete the cdrom stanzas. But that is your call.
If done correctly there will be no mirrorservice.org sections/stanzas.

Next two commands sets the desired mirror site, then removes the Debug Backport Testing stanzas.

urpmi.addmedia --distrib http://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64

urpmi.removemedia -y Debug Backport Testing

As a result each stanza will have a key-ids: line and all the *update* stanzas will have the update keyword. Example update stanza:

Core\ Updates http://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64/media/core/updates {
  key-ids: 80420f66
  update
}


I will then suggest that
  urpmi.update --wget  -a --wait-lock
should run without problems, assuming www.mirrorservice.org is not in the mist of pulling in updates from the Mageia mirror and you, your ISP and www.mirrorservice.org do not have any problems.

If no errors, run
  drakrpm-edit-media --expert
and verify all desired selections are checked in the Enabled column, and all desired *update* selections are checked under the Update column.

Thanks very much for those suggestions. I have done exactly what you said and the update proceeded without any problems.

So now my system is back and setup properly it is still not behaving as I'd expect - it takes a long time to work out what the installed packages are.

I just tried to install some packages and after about 5 minutes it crashed with a 'Bus error'. I don't think this should ever happen - I'd have thought 'Bus error' is always a bug.


[root@localhost rpm]# drakrpm
Ignore the following Glib::Object::Introspection & Gtk3 warnings
Subroutine Gtk3::main redefined at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 525.
Use of uninitialized value $value in numeric eq (==) at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 779.
Use of uninitialized value $value in numeric eq (==) at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 779.
Use of uninitialized value $value in numeric eq (==) at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 779.
Use of uninitialized value $value in numeric eq (==) at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 779.
Use of uninitialized value $value in numeric eq (==) at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 779.
Use of uninitialized value $value in numeric eq (==) at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 779.
Use of uninitialized value $value in numeric eq (==) at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 779.
Impossible to set by_group view as default
getting lock on urpmi
examining synthesis file [/var/lib/urpmi/Core Release2/synthesis.hdlist.cz]
examining synthesis file [/var/lib/urpmi/Core Updates/synthesis.hdlist.cz]
examining synthesis file [/var/lib/urpmi/Nonfree Release2/synthesis.hdlist.cz]
examining synthesis file [/var/lib/urpmi/Nonfree Updates/synthesis.hdlist.cz]
examining synthesis file [/var/lib/urpmi/Core 32bit Release/synthesis.hdlist.cz]
examining synthesis file [/var/lib/urpmi/Core 32bit Updates/synthesis.hdlist.cz]
examining synthesis file [/var/lib/urpmi/Nonfree 32bit Release/synthesis.hdlist.cz]
examining synthesis file [/var/lib/urpmi/Nonfree 32bit Updates/synthesis.hdlist.cz]
unlocking urpmi database
not using metalink since requested downloader does not handle it
retrieved ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64/media/core/release media_info/MD5SUM
retrieved ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64/media/core/release media_info/20170714-192548-info.xml.lzma
getting information from /var/lib/urpmi/Core Release2/info.xml.lzma
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
getting lock on urpmi
getting exclusive lock on rpm


retrieving rpm files from medium "Core Release2"...
    ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64/media/core/release/lyx-2.2.3-1.mga6.x86_64.rpm
retrieved ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64/media/core/release lyx-2.2.3-1.mga6.x86_64.rpm
...retrieving done
installing lyx-2.2.3-1.mga6.x86_64.rpm from /var/cache/urpmi/rpms
starting installing packages
created transaction for installing on / (remove=0, install=0, upgrade=1)
removing installed rpms (lyx-2.2.3-1.mga6.x86_64.rpm) from /var/cache/urpmi/rpms
unlocking urpmi database
unlocking rpm database
getting lock on urpmi
examining synthesis file [/var/lib/urpmi/Core Release2/synthesis.hdlist.cz]
examining synthesis file [/var/lib/urpmi/Core Updates/synthesis.hdlist.cz]
examining synthesis file [/var/lib/urpmi/Nonfree Release2/synthesis.hdlist.cz]
examining synthesis file [/var/lib/urpmi/Nonfree Updates/synthesis.hdlist.cz]
examining synthesis file [/var/lib/urpmi/Core 32bit Release/synthesis.hdlist.cz]
examining synthesis file [/var/lib/urpmi/Core 32bit Updates/synthesis.hdlist.cz]
examining synthesis file [/var/lib/urpmi/Nonfree 32bit Release/synthesis.hdlist.cz]
examining synthesis file [/var/lib/urpmi/Nonfree 32bit Updates/synthesis.hdlist.cz]
unlocking urpmi database
getting information from /var/lib/urpmi/Core Release2/info.xml.lzma
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.                                                    
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
getting information from /var/lib/urpmi/Core Release2/info.xml.lzma
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
getting information from /var/lib/urpmi/Core Release2/info.xml.lzma
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
getting information from /var/lib/urpmi/Core Release2/info.xml.lzma
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
getting information from /var/lib/urpmi/Core Release2/info.xml.lzma
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
getting information from /var/lib/urpmi/Core Release2/info.xml.lzma
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
getting information from /var/lib/urpmi/Core Release2/info.xml.lzma
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
retrieved ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64/media/core/updates media_info/MD5SUM
retrieved ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64/media/core/updates media_info/20171129-184801-info.xml.lzma
getting information from /var/lib/urpmi/Core Updates/info.xml.lzma
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
Argument "bold" isn't numeric in subroutine entry at /usr/lib/perl5/vendor_perl/5.22.3/Gtk3.pm line 1600.
getting lock on urpmi
getting exclusive lock on rpm


retrieving rpm files from medium "Core Release2"...
    ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64/media/core/release/kdiff3-0.9.98-8.mga6.x86_64.rpm
    ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64/media/core/release/lib64uv1-1.9.0-4.mga6.x86_64.rpm                             
    ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64/media/core/release/cmake-3.7.2-3.mga6.x86_64.rpm                                
    ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64/media/core/release/kdbg-2.5.4-5.mga6.x86_64.rpm                                 
    ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64/media/core/release/cppcheck-1.77-1.mga6.x86_64.rpm                              
    ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64/media/core/release/lib64tinyxml2_3-3.0.0-2.mga6.x86_64.rpm                      
    ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64/media/core/release/cmake-qtgui-3.7.2-3.mga6.x86_64.rpm                          
    ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64/media/core/release/lib64jsoncpp1-1.6.5-6.mga6.x86_64.rpm                        
retrieved ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64/media/core/release kdiff3-0.9.98-8.mga6.x86_64.rpm lib64uv1-1.9.0-4.mga6.x86_64.rpm cmake-3.7.2-3.mga6.x86_64.rpm kdbg-2.5.4-5.mga6.x86_64.rpm cppcheck-1.77-1.mga6.x86_64.rpm lib64tinyxml2_3-3.0.0-2.mga6.x86_64.rpm cmake-qtgui-3.7.2-3.mga6.x86_64.rpm lib64jsoncpp1-1.6.5-6.mga6.x86_64.rpm
...retrieving done
retrieving rpm files from medium "Core Updates"...
    ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64/media/core/updates/fossil-2.3-1.mga6.x86_64.rpm                                 
retrieved ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64/media/core/updates fossil-2.3-1.mga6.x86_64.rpm
...retrieving done
installing kdiff3-0.9.98-8.mga6.x86_64.rpm lib64uv1-1.9.0-4.mga6.x86_64.rpm cmake-3.7.2-3.mga6.x86_64.rpm kdbg-2.5.4-5.mga6.x86_64.rpm fossil-2.3-1.mga6.x86_64.rpm cppcheck-1.77-1.mga6.x86_64.rpm lib64tinyxml2_3-3.0.0-2.mga6.x86_64.rpm cmake-qtgui-3.7.2-3.mga6.x86_64.rpm lib64jsoncpp1-1.6.5-6.mga6.x86_64.rpm from /var/cache/urpmi/rpms
starting installing packages
created transaction for installing on / (remove=0, install=0, upgrade=9)
removing installed rpms (kdiff3-0.9.98-8.mga6.x86_64.rpm lib64uv1-1.9.0-4.mga6.x86_64.rpm cmake-3.7.2-3.mga6.x86_64.rpm kdbg-2.5.4-5.mga6.x86_64.rpm fossil-2.3-1.mga6.x86_64.rpm cppcheck-1.77-1.mga6.x86_64.rpm lib64tinyxml2_3-3.0.0-2.mga6.x86_64.rpm cmake-qtgui-3.7.2-3.mga6.x86_64.rpm lib64jsoncpp1-1.6.5-6.mga6.x86_64.rpm) from /var/cache/urpmi/rpms
unlocking urpmi database
unlocking rpm database
getting lock on urpmi
examining synthesis file [/var/lib/urpmi/Core Release2/synthesis.hdlist.cz]
examining synthesis file [/var/lib/urpmi/Core Updates/synthesis.hdlist.cz]
examining synthesis file [/var/lib/urpmi/Nonfree Release2/synthesis.hdlist.cz]
examining synthesis file [/var/lib/urpmi/Nonfree Updates/synthesis.hdlist.cz]
examining synthesis file [/var/lib/urpmi/Core 32bit Release/synthesis.hdlist.cz]
examining synthesis file [/var/lib/urpmi/Core 32bit Updates/synthesis.hdlist.cz]
examining synthesis file [/var/lib/urpmi/Nonfree 32bit Release/synthesis.hdlist.cz]
examining synthesis file [/var/lib/urpmi/Nonfree 32bit Updates/synthesis.hdlist.cz]
Bus error (core dumped)
[root@localhost rpm]#

(In reply to Simon Geard from comment #17)
> Thanks very much for those suggestions. I have done exactly what you said
> and the update proceeded without any problems.
> 
> So now my system is back and setup properly it is still not behaving as I'd
> expect - it takes a long time to work out what the installed packages are.
> 
> I just tried to install some packages and after about 5 minutes

That does not surprise me, doing install/update rpms without having all media selections/stanzas working has caused problems that the rpm/urpmi can not resolve. I learned the hard way thinking it would not matter.  :(

>  it crashed with a 'Bus error'. I don't think this should ever happen
> - I'd have thought 'Bus error' is always a bug.

Yup, that is always a bug report from me, IF I can create a test case to reproduce the problem.

I am going to ignore all the perl errors since it seems that you did not launch rpmdrak via mcc. I have already written up bug reports for the perl errors seen when launching mcc.

> retrieving rpm files from medium "Core Release2"...
>    
> ftp://www.mirrorservice.org/sites/mageia.org/pub/mageia/distrib/6/x86_64/
> media/core/release/lyx-2.2.3-1.mga6.x86_64.rpm
> retrieved

You are doing an update so I did not expect rpms from non /update { stanzas.

I hope you did not place a check mark in the Update column for selections which did not have update in their name. 

For instance, if no check mark in that column for "Core Release2" then based on the lyx-2.2.3-1.mga6.x86_64.rpm being pulled in, I am pretty sure you will wind up having to do a clean install, then updates to get your system to a stable condition, rpm wise. 

Having screwed up my installs a few times, 99% were Cauldron and not Official like yours, I can only recommend a clean install using a known mirror.

I scripted setting the mirror, verify media is up to date, pull down updates, then install updates time after time.  Makes everything go much faster and are pron to far less problems that you have shown so far except, for this last batch. You have all the basic commands for doing so in this thread.

I haven't seen any similar reports that can be counted as confirmation that there is an unknown bug in our tools that contributed to the problems here.

Apart from known bugs, about us *really* needing mirrorbrain for better mirror selection, see bug #3166, and about the not-hidden warnings that rpmdrake tells us to ignore ("Ignore the following Glib::Object::Introspection & Gtk3 warnings"), I don't see anything that could be fixed on our side.

Besides, there was no action in this report since over half a year ago.

Closing as OLD

Resolution: (none) => OLD
Status: UNCONFIRMED => RESOLVED