[rhelv6-list] why do errata and package updates fail?

Tue Oct 9 18:31:39 UTC 2012

On 10/09/2012 12:33 PM, Mirko Vukovic wrote:
> On Tue, Oct 9, 2012 at 12:18 PM, thomas at redhat.com <thomas at redhat.com> wrote:
>> On 10/09/2012 11:10 AM, Mirko Vukovic wrote:
>>>
>>> yum update installed 26 packages.  The system shows as up-to-date on RHN.
>>
>>
>> Without seeing the logs from the previous commands, it's almost impossible
>> to tell what happened.
>>
> ...
>> What repositories do you have enabled? What is the output of the following
>> command?
>>
>> yum -v repolist
>
> Here is output of `yum -v repolist'
>
> Loading "rhnplugin" plugin
> Loading "product-id" plugin
> Loading "refresh-packagekit" plugin
> Loading "security" plugin
> Loading "subscription-manager" plugin
> Updating certificate-based repositories.
> Unable to read consumer identity
> Config time: 0.109
> Looking for repo options for [main]
> Looking for repo options for [rhel-x86_64-workstation-6]
> Repo 'rhel-x86_64-workstation-6' setting option 'enabled' = '1'
> Repo 'rhel-x86_64-workstation-6' setting option 'gpgcheck' = '1'
> Looking for repo options for [rhel-x86_64-workstation-supplementary-6]
> Repo 'rhel-x86_64-workstation-supplementary-6' setting option 'enabled' = '1'
> Repo 'rhel-x86_64-workstation-supplementary-6' setting option 'gpgcheck' = '1'
> Looking for repo options for [rhel-x86_64-workstation-optional-6]
> Repo 'rhel-x86_64-workstation-optional-6' setting option 'enabled' = '1'
> Repo 'rhel-x86_64-workstation-optional-6' setting option 'gpgcheck' = '1'
> Yum Version: 3.2.29
> Setting up Package Sacks
> pkgsack time: 0.055
> Repo-id      : elrepo
> Repo-name    : ELRepo.org Community Enterprise Linux Repository - el6
> Repo-updated : Mon Oct  8 01:31:18 2012
> Repo-pkgs    : 219
> Repo-size    : 931 M
> Repo-baseurl : http://elrepo.org/linux/elrepo/el6/x86_64/
> Repo-mirrors : http://elrepo.org/mirrors-elrepo.el6
> Repo-expire  : 21,600 second(s) (last: Tue Oct  9 11:55:29 2012)
>
> Repo-id      : epel
> Repo-name    : Extra Packages for Enterprise Linux 6 - x86_64
> Repo-revision: 1349708267
> Repo-tags    : binary-x86_64
> Repo-updated : Mon Oct  8 10:59:47 2012
> Repo-pkgs    : 7,887
> Repo-size    : 6.9 G
> Repo-metalink: https://mirrors.fedoraproject.org/metalink?repo=epel-6&arch=x86_64
>    Updated    : Mon Oct  8 10:59:47 2012
> Repo-baseurl : http://fedora-epel.mirror.lstn.net/6/x86_64/ (38 more)
> Repo-expire  : 21,600 second(s) (last: Tue Oct  9 11:55:31 2012)
>
> Repo-id      : rhel-x86_64-workstation-6
> Repo-name    : Red Hat Enterprise Linux Workstation (v. 6 for x86_64)
> Repo-updated : Mon Oct  8 08:56:32 2012
> Repo-pkgs    : 9,349
> Repo-size    : 15 G
> Repo-baseurl : https://xmlrpc.rhn.redhat.com/XMLRPC/GET-REQ/rhel-x86_64-workstation-6
> Repo-expire  : 21,600 second(s) (last: Tue Oct  9 11:55:32 2012)
>
> Repo-id      : rhel-x86_64-workstation-optional-6
> Repo-name    : RHEL Workstation Optional (v. 6 for x86_64)
> Repo-updated : Mon Oct  8 08:56:32 2012
> Repo-pkgs    : 4,565
> Repo-size    : 3.3 G
> Repo-baseurl : https://xmlrpc.rhn.redhat.com/XMLRPC/GET-REQ/rhel-x86_64-workstation-optional-6
> Repo-expire  : 21,600 second(s) (last: Tue Oct  9 11:55:33 2012)
>
> Repo-id      : rhel-x86_64-workstation-supplementary-6
> Repo-name    : RHEL Workstation Supplementary (v. 6 for x86_64)
> Repo-updated : Tue Oct  9 03:53:35 2012
> Repo-pkgs    : 276
> Repo-size    : 3.6 G
> Repo-baseurl : https://xmlrpc.rhn.redhat.com/XMLRPC/GET-REQ/rhel-x86_64-workstation-supplementary-6
> Repo-expire  : 21,600 second(s) (last: Tue Oct  9 11:55:34 2012)

Not a huge deal, but I've seen some really weird issues with third party 
repositories like ELRepo and even EPEL. Don't get me wrong, I use EPEL 
and other third party repositories like ATRPMs, but sometimes it's 
just... wonky.

>> I almost guarantee that the answers are in the system event log. Go to RHN,
>> click on the system, choose "Events" in the menu just below the hostname,
>> then choose "History" to see what happened.
>>
>
> Thanks for showing me how to get to history.  This is what I found:
> - Some were failures because of existing lock on var/run/yum.pid
> - Other were package dependency failures: u'Protected multilib
> versions: ...x86_64 != ...i686 (code 18)
>
> I checked (ps -A) and there is currently no yum running.

I've seen both of those before... If you happen to run a yum command 
while PackageKit or yum-updatesd are running a job, your yum command 
gets blocked. It stinks, but generally you can just wait a while and run 
your yum command again.

The Protected multilib error tells me you might have two versions (i686 
and x86_64) of a package installed. This is not a big deal, but the 
nature of using RHN to do updates is that it processes each errata and 
the associated RPMs serially. So if you have one erratum which addresses 
the 32-bit version of a file and another which addresses the 64-bit 
version, you might see that. You might get an error when it tries to 
update the first package because of common dependencies with the second. 
Without more of the error message, it's hard to tell, but I don't see 
anything which is seriously alarming in what you've posted. Frustrating, 
for sure. But not unexpected.

Might not be a bad idea to open a ticket with support... they are the 
best way to get issues like this in front of our developers (this list 
is NOT - some developers watch it, but there is no service level 
agreement around mailing lists).
-- 
Thomas Cameron, RHCA, RHCSS, RHCDS, RHCVA, RHCX
Chief Architect, Central US
512-241-0774 office / 512-585-5631 cell
http://people.redhat.com/tcameron/
IRC: choirboy / AIM: rhelguy / Yahoo: rhce_guy

Red Hat named to Forbes 2012 World's Most Innovative Companies list:
http://www.redhat.com/innovation