RE: Local Yum Repo HowTo On Line
by Timothy A. Holmes
> > > > > On Tue, 2005-06-28 at 21:06 -0400, David Cary Hart wrote:
> > > > > > This is probably a worthwhile endeavor if you have three or
more
> > > > > > machines to keep updated. There are a number of these
floating
> around. I
> > > > > > have tried to update and simplify the process.
> > > > > >
> > > > > > Let me know what you think.
> > > > > > http://www.TQMcube.com/repo.htm
> > > >
> > > > Very good! However, why the need to copy all the CD/DVD onto the
HD?
> Why
> > > > not grab the files from Red Hat or a mirror? Have I missed
> something?
> > >
> > > That would work equally well but David is assuming (rightly) that
most
> > > people will already have the CD/DVD (or at least the ISOs, which
are
> > > equivalent) from which they've done the OS install, so why
download
> > > everything again?
> >
> > Sorry Paul!
> >
> > I meant the files createrepo /var/www/html/yum/base created. From
the
> > repo.htm this is what I assumed was the purpose of copying the files
> > from the CD/DVD. Hence, have I missed something? While Red Hat 6.2
was
> > supported, I just ftp'd the updates into a directory "updates".
> > Incidentally, I have two boxes, one an Intel and the other an Alpha.
>
> Ah, I see. It doesn't really matter where you get the files from as
long
> as you maintain their relationship to each other directory-wise. For
> instance, on download.fedora.redhat.com, the structure is:
>
> Fedora -> RPMS -> all-rpm-files
> |
> repodata
>
> So to get from the repodata directory to the actual packages you
> go ../Fedora/RPMS
>
> You'd need to maintain this relationship if you just copied the files
> from there, so you'd have to have something like:
>
> /var/www/html/yum/base/Fedora/RPMS/*.rpm
> /var/www/html/yum/base/repodata
>
> Using createrepo locally allows you to save a level of hierarchy and
is
> actually probably faster than downloading the metadata anyway.
>
> Note that for the "updates" area, the document does suggest copying
the
> metadata along with the packages, just as you did for Red Hat 6.2.
>
> Paul.
> --
> Paul Howarth <paul(a)city-fan.org>
[Tim Holmes]
Good Morning -- This is a fascinating thread, as I am planning to
establish an FC3 Repo in the next week or two. There have been some
significant path changes etc discussed, and I was wondering if the
originator of the thread could post an updated procedure for us newer
users to follow?
If I can be of assistance in helping to get this done, please let me
know
TIM
18 years, 9 months
Yum dependency problem with openoffice
by Harald Grossauer
Hi,
running "yum update" recently gives me:
--> Running transaction check
--> Processing Dependency: openoffice.org-core = 1:1.9.104-2 for
package: openoffice.org-pyuno
--> Finished Dependency Resolution
Error: Missing Dependency: openoffice.org-core = 1:1.9.104-2 is needed
by package openoffice.org-pyuno
Can it be that the openoffice.org-core package is missing in the repos?
18 years, 9 months
[FC4] jpackage.org java build fails
by Christoph Wickert
Hi
I'm trying to build a custom java rpm from jpackage.org.
> [chris@hal9000 SPECS]$ rpmbuild -ba java-1.5.0-sun.spec
> Executing(%prep): /bin/sh -e /var/tmp/rpm-tmp.41418
> + umask 022
> + cd /home/chris/develope/rpmbuild/BUILD
> + LANG=C
> + export LANG
> + unset DISPLAY
[...]
> + /usr/lib/rpm/check-rpaths /usr/lib/rpm/check-buildroot
> ERROR: file '/usr/share/java-1.5.0-sun/demo/jvmti/mtrace/lib/libmtrace_g.so' contains an invalid rpath '$ORIGIN/../../../../lib/i386' in [$ORIGIN/../../../../lib/i386:$ORIGIN/../../../../jre/lib/i386]
> ERROR: file '/usr/share/java-1.5.0-sun/demo/jvmti/mtrace/lib/libmtrace_g.so' contains an invalid rpath '$ORIGIN/../../../../jre/lib/i386' in [$ORIGIN/../../../../lib/i386:$ORIGIN/../../../../jre/lib/i386]
> ERROR: file '/usr/share/java-1.5.0-sun/demo/jvmti/mtrace/lib/libmtrace.so' contains an invalid rpath '$ORIGIN/../../../../lib/i386' in [$ORIGIN/../../../../lib/i386:$ORIGIN/../../../../jre/lib/i386]
> ERROR: file '/usr/share/java-1.5.0-sun/demo/jvmti/mtrace/lib/libmtrace.so' contains an invalid rpath '$ORIGIN/../../../../jre/lib/i386' in [$ORIGIN/../../../../lib/i386:$ORIGIN/../../../../jre/lib/i386]
> ERROR: file '/usr/share/java-1.5.0-sun/demo/jvmti/heapTracker/lib/libheapTracker.so' contains an invalid rpath '$ORIGIN/../../../../lib/i386' in [$ORIGIN/../../../../lib/i386:$ORIGIN/../../../../jre/lib/i386]
> ERROR: file '/usr/share/java-1.5.0-sun/demo/jvmti/heapTracker/lib/libheapTracker.so' contains an invalid rpath '$ORIGIN/../../../../jre/lib/i386' in [$ORIGIN/../../../../lib/i386:$ORIGIN/../../../../jre/lib/i386]
> ERROR: file '/usr/share/java-1.5.0-sun/demo/jvmti/heapTracker/lib/libheapTracker_g.so' contains an invalid rpath '$ORIGIN/../../../../lib/i386' in [$ORIGIN/../../../../lib/i386:$ORIGIN/../../../../jre/lib/i386]
> ERROR: file '/usr/share/java-1.5.0-sun/demo/jvmti/heapTracker/lib/libheapTracker_g.so' contains an invalid rpath '$ORIGIN/../../../../jre/lib/i386' in [$ORIGIN/../../../../lib/i386:$ORIGIN/../../../../jre/lib/i386]
> Fehler: Bad exit status from /var/tmp/rpm-tmp.1543 (%install)
>
The directory /usr/share/java-1.5.0-sun does not exist, I only have
an empty /usr/share/java-1.5.0 which was created by
jpackage-utils-1.6.3-1jpp_1rh. What am I missing?
Christoph
18 years, 9 months
FC4 nfs install failure
by Paul Elliott
FC4 is failing to network install with nfs!
I get "that directory could not be mounted from the server"
error message.
Same server works fine installing FC3 same setup!
Server /var/log/messages indicate no attempt to nfs
mount the directory.
Any ideas?
--
Paul Elliott 1(512)837-1096
pelliott(a)io.com PMB 181, 11900 Metric Blvd Suite J
http://www.io.com/~pelliott/pme/ Austin TX 78758-3117
18 years, 9 months
problems after updating selinux-policy-targeted
by Martin Marques
After the update I wasn't able to login through ssh. Restarted sshd and
everything started working OK.
Now I have this problem when login in with KDE (from the X console). I
restarted KDM (killall -HUP X) but it didn't work.
These are the error logs I get:
Jun 29 10:30:08 mail kernel: audit(1120051808.126:0): avc: denied
{ execmod } for pid=2952 comm=prefdm path=/lib/libdl-2.3.5.so dev=hda1
ino=50817 scontext=user_u:system_r:unconfined_t tcontext=root:object_r:lib_t
tclass=file
Jun 29 10:30:09 mail kernel: audit(1120051809.938:0): avc: denied
{ execmod } for pid=2995 comm=prefdm path=/lib/libdl-2.3.5.so dev=hda1
ino=50817 scontext=user_u:system_r:unconfined_t tcontext=root:object_r:lib_t
tclass=file
Jun 29 10:30:11 mail kernel: audit(1120051811.282:0): avc: denied
{ execmod } for pid=3044 comm=prefdm path=/lib/libdl-2.3.5.so dev=hda1
ino=50817 scontext=user_u:system_r:unconfined_t tcontext=root:object_r:lib_t
tclass=file
Jun 29 10:30:12 mail kernel: audit(1120051811.670:0): avc: denied
{ execmod } for pid=3099 comm=prefdm path=/lib/libdl-2.3.5.so dev=hda1
ino=50817 scontext=user_u:system_r:unconfined_t tcontext=root:object_r:lib_t
tclass=file
Jun 29 10:30:13 mail kernel: audit(1120051813.903:0): avc: denied
{ execmod } for pid=3143 comm=prefdm path=/lib/libdl-2.3.5.so dev=hda1
ino=50817 scontext=user_u:system_r:unconfined_t tcontext=root:object_r:lib_t
tclass=file
Jun 29 10:30:14 mail kernel: audit(1120051814.889:0): avc: denied
{ execmod } for pid=3194 comm=prefdm path=/lib/libdl-2.3.5.so dev=hda1
ino=50817 scontext=user_u:system_r:unconfined_t tcontext=root:object_r:lib_t
tclass=file
Jun 29 10:30:16 mail kernel: audit(1120051816.129:0): avc: denied
{ execmod } for pid=3238 comm=prefdm path=/lib/libdl-2.3.5.so dev=hda1
ino=50817 scontext=user_u:system_r:unconfined_t tcontext=root:object_r:lib_t
tclass=file
Jun 29 10:30:17 mail kernel: audit(1120051817.941:0): avc: denied
{ execmod } for pid=3282 comm=prefdm path=/lib/libdl-2.3.5.so dev=hda1
ino=50817 scontext=user_u:system_r:unconfined_t tcontext=root:object_r:lib_t
tclass=file
Jun 29 10:30:18 mail kernel: audit(1120051818.943:0): avc: denied
{ execmod } for pid=3327 comm=prefdm path=/lib/libdl-2.3.5.so dev=hda1
ino=50817 scontext=user_u:system_r:unconfined_t tcontext=root:object_r:lib_t
tclass=file
Jun 29 10:30:20 mail kernel: audit(1120051820.651:0): avc: denied
{ execmod } for pid=3373 comm=prefdm path=/lib/libdl-2.3.5.so dev=hda1
ino=50817 scontext=user_u:system_r:unconfined_t tcontext=root:object_r:lib_t
tclass=file
Jun 29 10:30:21 mail init: Id "x" respawning too fast: disabled for 5 minutes
--
select 'mmarques' || '@' || 'unl.edu.ar' AS email;
---------------------------------------------------------
Martín Marqués | Programador, DBA
Centro de Telemática | Administrador
Universidad Nacional
del Litoral
---------------------------------------------------------
18 years, 9 months
K-programs font problem
by Dotan Cohen
In my new FC4 installation (not upgrade) I have squares appearing in
Konquerer in places where the source code specifies emplty spaces:
. Also, in Kmail I cannpt read hebrew mails. I tried all the
different hebrew encodings, but none of them succeeded in rendering
the message. I think that if I solve the first problem, then it will
also fix the second, because in the default (automatic) encoding I see
the same squares that I see in Konquerer.
Dotan Cohen
http://lyricslist.com/lyrics/artist_albums/332/mccartney_paul.php
McCartney, Paul Song Lyrics
18 years, 9 months
Linux installs on CF Flash chips.
by Jim Cornette
Having an installation of Linux on a CF Flash, I am wondering if it is
needed to check for bad blocks during file formatting. I am also
wondering if using tune2fs -c 0 and tune2fs -i 0 is needed or is a bad idea.
I did not set up this CF installation, but am curious as to CF Flashes
and their reliability.
The options they used were for tune2fs and formatting the filesystem with
mkfs.ext2 -j /dev/hda
Would passing the -c flag to mkfs.ext2 along with the -j option for
journalling be needed for flashes? Will it mark the blocks as bad before
the filesystem is used for anything?
Jim
--
A child of five could understand this! Fetch me a child of five.
18 years, 9 months
Thank you Fedora!
by Steve Croteau
I just wanted to let all of you know how much I appreciate your effort in
developing, contributing and caring for this project. I'm currently using
FC3 on two legacy boxes (AMD 450 & Intel 466) in my home office and I see no
reason to own MS software.
I'm not bashing here, just stating the facts as I see them. My computers may
be slow but I choose not to play the constant hardware upgrade game that most
in the MS world play due to a bloated OS requiring more & more system
resources. Linux affords me that choice. The ROI these two computers have
generated would be hard to measure.
Being a small business owner, Linux makes sense!
Thank you,
Steve Croteau
Serpent Cymbals
SerpentCymbals(a)comcast.net
18 years, 9 months
Some thoughts for the future
by Richard Kelsch
Since Fedora is supposed to be "bleeding-edge," I have thought of a few
things that may be cool for the future, not just for Fedora but Linux in
general:
* How about the ability to install via SRPMS for the specific CPU
platform I'm using? In other words, as an advanced install option,
allow me to have the installer use the SRPMS instead of the precompiled
RPMS, and allow me to define the compiler optimizations for CPU etc.
That would be cool. Ok, a slow install, but this isn't a production OS,
so why not? I hate having an Athlon-XP running code optimized for a
P4. Sure, an install could take days, but it would be an option I'd
use. Why? Because I can and it's certainly bleeding-edge.
* How about the ability to clone the configured and working version on
my hardrive to be burned on a DVD (dual layer and blue beam makes this
possible) for easy portability?
* Someone needs to make X more display hardware friendly.
* Gnome and KDE are nice and such, but even Apple knew how to exploit
the display hardware to make a cool interface. So much so that MS is
trying to copy and "improve" it for their next release. There's more to
a gui now days than skinning. Fedora needs a cooler looking GUI.
* Better documentation of changes that could result in problems with
legacy software, and techniques on how to fix them, instead of the usual
"we took [this] out" without further explanation.
Can you immagine a day when the Fedora List has nothing but praise
instead of a plethora of problems? One can dream....
Rich
18 years, 9 months