[Spacewalk-list] strange thing with a clean spacewalk install

[Michiel van Es]

-------- Original Message --------
Subject: Re: [Spacewalk-list] strange thing with a clean spacewalk install
From: Justin Sherrill <jsherril at redhat.com>
To: spacewalk-list at redhat.com <spacewalk-list at redhat.com>
Date: 01/13/2010 07:46 PM

> On 1/13/10 9:38 AM, Michiel van Es wrote:
>> Hi,

Hi Justin,

>>
>> I have some strange behaviour with Spacewalk 0.7 on CentOS 5 (i386) and
>> CentOS 5 clients (i386 and x86_64).
>>
>> For example: I see there is a new kernel downloaded and available in
>> spacewalk: 2.6.18-164.10.1.el5
>>
>> But it is not provided as upgrade to the new systems, an older version
>> is provided:2.6.18-164.10.1.el5
>>
>> Why is spacewalk installing the old version first and not directly the
>> newer version? And why doesn't the new kernel version pop up as upgrade
>> at the clients upgrades packages?
>
> Sounds like your yum cache is out of date.  Is taskomatic running?
> (service taskomatic status).
>
> Try deleting /var/cache/rhn/repodata/CHANNEL_LABEL  and then try running
> 'yum update'.  It'll get an error (404) but that should initiate a
> rebuild of your yum repodata.  You can monitor that directory to see it
> complete. (May take up to an hour).

I found that manually running rhn-sync-profile does the trick.
The problem is that some machines are automatically syncing there 
packages updates to spacewalk en some don't.

>
>>
>>
>> Other thing I've got, while :
>>
>> 1-25 of 31 (48 selected)
>>
>> I don't get the numbers spacewalk gave me, why is it saying I have 1-25
>> of 31 machines to select (and it says 48 selected) ??
>
> This is odd.  Try clicking 'unselect all' and see if the problem occurs
> again.

Hmm I also got it by selecting packages.
If you wish I can send some screenshots?

>
> -Justin

Michiel
>
>
>>
>>
>
>