[Spacewalk-list] Force install, still looking for any answers

m.roth2006 at rcn.com m.roth2006 at rcn.com
Tue Apr 7 21:15:02 UTC 2009


Cliff,

>Date: Tue, 07 Apr 2009 16:05:51 -0400
>From: Cliff <cperry at redhat.com>  
>m.roth2006 at rcn.com wrote:
>>> Date: Tue, 7 Apr 2009 13:58:53 -0400
>>> From: Jesus Rodriguez <jesusr at redhat.com>  
>>> On Tue, Apr 07, 2009 at 01:05:54PM -0400, m.roth2006 at rcn.com wrote:
>>>> Asking again, since I got no responses:
>>>>
>>>> I created a new CentOS 5.3 repository Friday. Today, I unsubscribed my test system to 5.2, and subscribed it to 5.3 and its child channels. I tried to upgrade. 
>> <snip>
>>> What I got was that it had failed, because of rpm conflicts.
<snip>
>>> And what failed? errors would be helpful in debugging your problem.
>> 
>> As I said, above: the errors listed for the system failure were that a bunch of packages conflicted with existing packages or dependencies. Now, I thought I was doing an upgrade, and that should have not had problem, but this is what I got. 
>
>So, what I'm asking is whether there's some way to tell spacewalk, when 
>it does the push, use --force --nodeps.
>
>
>I *assume* 'push' - you mean, the clients pulling the CentOS 5.3 content 
>off your spacewalk channels?

Ok, am I *that* obtuse? No. On the spacewalk server, I scheduled a push of an upgrade, 143 packages including the kernel, to UPGRADE my test system to CentOS 5.3 from 5.2. The scheduled *PUSH* failed, telling me (in schedules, failed system), 
Error while executing packages action: Transaction Check Error: file /etc/ldap.conf from install of nss_ldap-253-17.el5 conflicts with file from package nss_ldap-253-13.el5_2.1 <snippage of another bunch of the same kind of failures>

       mark




More information about the Spacewalk-list mailing list