[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

RE: [rhn-users] up2date tcpdump-3.6.3-3:14 problem


Thanks for all that great info and effort!  Fortunately I was able to update
everything just fine yesterday and server so far is running like a champ.
I'll make sure to read all of your info.


-----Original Message-----
From: rhn-users-admin redhat com [mailto:rhn-users-admin redhat com]On
Behalf Of Mike Chambers
Sent: Thursday, May 01, 2003 10:34 AM
To: RedHat - RHN
Subject: Re: [rhn-users] up2date tcpdump-3.6.3-3:14 problem

On Tue, 2003-04-29 at 23:45, DB wrote:

> I'm trying to update this package manually.  I ran the up2date
> tcpdump-3.6.3-3:14 command and got the below response, even though RHN
> this package needs to be updated.  Did I miss a step?

I know you already did all these updates, by reading the whole thread (I
was gone for few days out of town), so this is a tad late.

1st issue - When you want to upgrade or install (yes, up2date package
will install it if you don't already have it), just use the package
name, no version numbers need to be included.  Example, up2date
tcpdump.  The only time version numbers are needed are usually when
removing a package that has more than one version installed, such as the

2nd issue - You discussed with others on being *at* the server location,
or being remote, when upgrading using up2date.  There is no difference
for either one, except that if something goes wrong during boot up or
such, you won't see any messages if it hangs, so therefore requiring you
be there to see what is going on, and manipulating the keyboard to get
it to boot and fix it if needed.

3rd issue - When upgrading the kernel, it uses the command of rpm -ivh
kernel, which installs the new kernel, and leaves the old one.  It also
modifies the grub.conf file to boot to the newer kernel by default, but
with the old kernel still installed, you can have it boot to the older
one in case of problems or whatever.  Others mentioned about the size of
the /boot partition and the stability of the new kernel before removing
the old one.  It requires certain amount of space to install a new
kernel, so if for example, you had 5 different kernel updates installed
and never remove done, your /boot partition might be getting full (IF
/boot is a seperate partition, otherwise it just falls under /) and you
would then need to remove older kernels that are no longer needed.  BTW,
that is when you need the version number when using rpm.  If you have 3
kernels installed, kernel-1.2.3, kernel-2.4.5 and kernel-6.7.8, you
could rpm -e kernel-1.2.3 and that will remove the older kernel and free
up some space.  You could even rpm -e kernel-2.4.5 and remove that one
as well if you have been using the new kernel for a while and no

4th issue - Your profiles on RHN.  What *I* do, is when I upgrade (or in
my case, a fresh install) to a new release, I just remove the existing
machine profile all together, then register it when it's done
installing, so I don't run into any problems.  I then configure up2date
to skip/not skip packages, and other things, so I don't forget and it's
setup how I like it and it's ready to go.

Hope this helps any and good luck on using RHN,

Mike Chambers
Madisonville, KY

"Hand over the code, and no one gets hurt!"

rhn-users mailing list
rhn-users redhat com

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]