[Yum] Downloading header problems

Troy Dawson dawson at fnal.gov
Tue May 24 15:09:27 UTC 2005


Howdy,
I'm not thinking this is a yum problem, probrubly a type of 
configuration problem, but I just can't figure it out.

Background
Release: Scientific Linux 4.0
yum-2.2.1-1.SL.noarch
rpm-4.3.3-7_nonptl.i386
python-2.3.4-14.1.i386

This is not a modified yum like our older one.  The only thing we do is 
pull the configuration files out and put them in a separate rpm.  So 
this is a real 2.2.1.

Problem
For a user in Australia, when he does an update for new security 
releases, instead of just pulling down a small header, it pulls down the 
full rpm, which takes quite a while.

So instead of
------------------------------
#yum update firefox

...snip...

Resolving Dependencies
--> Populating transaction set with selected packages. Please wait.
---> Downloading header for firefox to pack into transaction set.
firefox-1.0.4-1.4.1.i386. 100% |=========================|  66 kB    00:00
---> Package firefox.i386 0:1.0.4-1.4.1 set to be updated
--> Running transaction check

Dependencies Resolved
Transaction Listing:
   Update: firefox.i386 0:1.0.4-1.4.1 - sl-errata
Total download size: 16 M
---------------------------------

He is getting

---------------------------------
# yum update firefox

...snip...

Resolving Dependencies
--> Populating transaction set with selected packages. Please wait.
---> Downloading header for firefox to pack into transaction set.
firefox-1.0.4-1.4.1.i386. 100% |=========================|  16 MB    08:43
---> Package firefox.i386 0:1.0.4-1.4.1 set to be updated
--> Running transaction check

Dependencies Resolved
Transaction Listing:
   Update: firefox.i386 0:1.0.4-1.4.1 - sl-errata
Total download size: 16 M
-------------------------------------

As you can see, instead of downloading a 66 kB file, it is downloading a 
16 MB file, which is the same size as the rpm.

His config file, is the same config file as a standard except for one 
thing.  His cache line says
cachedir=/mnt/software/mirror/sl/40
which is a nfs mounted directory.  This is his master machine doing the 
yum'ing, and all the others have this mounted and do a 'yum -C ...' so 
they don't have to do the downloading.  But I really don't feel that 
it's the cache doing anything.

The server machine is the normal server machine, so his repository file 
looks like this
------------------
# cat /etc/yum.repos.d/sl-errata.repo
[main]
[sl-errata]
name=SL 4 base
baseurl=ftp://ftp.scientificlinux.org/linux/scientific/40/$basearch/errata/SL/RPMS/
enabled=1
-------------------
His yum.conf looks like this
-------------------
# cat /etc/yum.conf
[main]
cachedir=/mnt/software/mirror/sl/40
debuglevel=2
logfile=/var/log/yum.log
pkgpolicy=newest
distroverpkg=redhat-release
tolerant=1
exactarch=1
obsoletes=1
gpgcheck=1

# PUT YOUR REPOS HERE OR IN separate files named file.repo
# in /etc/yum.repos.d
-------------------

Sorry for making this so long, but I've been trying to figure this out, 
so I have alot of info. :)
Any hints as to where the problem lies?  Why he would be getting the 
full file instead of just a header?

Thanks
Troy Dawson
-- 
__________________________________________________
Troy Dawson  dawson at fnal.gov  (630)840-6468
Fermilab  ComputingDivision/CSS  CSI Group
__________________________________________________



More information about the Yum mailing list