On Tue, 19 Aug 2008, FRANCHISSEUR Robert wrote:
Hello,
I have the following problem with yum and openssh.
If I remove openssh and re-install it I got:
openssh-3.9p1-8.RHEL4.24 from sl-errata
but after 5 days it is automaticaly updated by yum
nightly updater to:
openssh-3.9p1-8.SL.4.22 from sl-base
This is from the sl-contrib repo, I don't understand why it is showing up
as sl-base...
I don't understand from where does this release comme from as in
sl-base (ftp.scientificlinux.org/linux/scientific/45/i386/SL/RPMS/) I see :
[ ] openssh-3.9p1-8.RHEL4.20.i386.rpm 03-May-2007 16:55 322K
furthermore :
[EMAIL PROTECTED] ~]# yum info openssh
Loading "kernel-module" plugin
Setting up repositories
Reading repository metadata in from local files
Installed Packages
Name : openssh
Arch : x86_64
Version: 3.9p1
Release: 8.SL.4.22
Size : 950 k
Repo : installed
Summary: The OpenSSH implementation of SSH protocol versions 1 and 2.
Description:
<snip>
Available Packages
Name : openssh
Arch : x86_64
Version: 3.9p1
Release: 8.RHEL4.24
Size : 347 k
Repo : sl-errata
Summary: The OpenSSH implementation of SSH protocol versions 1 and 2.
Description:
<snip>
Any explanation ?
Thanks for your help.
What does running:
yum -d 1 list openssh
show? For me on sl4x it shows:
Installed Packages
openssh.x86_64 3.9p1-8.RHEL4.24 installed
but if I enable sl-contrib I also get the sl one, e.g.
$ yum --enablerepo=sl-contrib -d 1 list openssh
Installed Packages
openssh.x86_64 3.9p1-8.RHEL4.24 installed
Available Packages
openssh.x86_64 3.9p1-8.SL.4.22 sl-contrib
Of course as far as rpm/yum are concerned 'SL' counts as newer than 'RHEL'
so if you have sl-contrib enabled it will be preferred and so count as a
newer update - even it it isn't really newer.
For reasons very like this in the scripts we run to do yum updates we have
stuff like:
...
$yumopts .= ' --disablerepo=\'*\' --enablerepo='.$stdrepos;
...
system ("yum $yumverb -y $yumopts update");
...
so we don't get surprised if a repo got turned on - usually just to test
something, we rarely *intentionally* permanently enable any other than the
ones we consider 'standard'.
--
Jon Peatfield, Computer Officer, DAMTP, University of Cambridge
Mail: [EMAIL PROTECTED] Web: http://www.damtp.cam.ac.uk/