On 10/09/2011 07:17 PM, Balcaen John wrote:
Le dimanche 9 octobre 2011 19:10:49 Frank Griffin a écrit :
I have two virtually identical cauldron systems which update from the
same local cauldron tree. One consistently gets messages during "urpmi
--auto-update" like "adding 16777 new rpms not available in existing
hdlist" which holds up the urpmi processing for literally minutes, and
one does not.
Does anyone know what this message indicates, what processing is
occurring during the long delay, and any reason why it should occur on
one system but not the other ?
You're probably using a local media without a hdlist inside.
I don't see how. Both systems install using HTTP from the same distro
tree and use the same script to delete and recreate media from that tree
after install:
********************************************************
[ftg@ftgme2 bin]$ cat customizeUrpmi.sh
#!/bin/sh
#
# Create urpmi sources
#
echo "Deleting existing urpmi sources"
urpmi.removemedia -a
echo "Creating urpmi sources"
urpmi.addmedia --no-md5sum --probe-hdlist --distrib
file:///mnt/cauldron/$HOSTTYPE
urpmi.update --no-ignore "Core Release Debug" "Core 32bit Release Debug"
"Nonfree Release Debug" "Nonfree Release" "Tainted Release"
urpmi.update -a
********************************************************
The urpmi.cfg files are virtually identical:
***************************************
[root@ftglap urpmi]# cmp -b /etc/urpmi/urpmi.cfg
/mnt/ftgme2/etc/urpmi/urpmi.cfg
/etc/urpmi/urpmi.cfg /mnt/ftgme2/etc/urpmi/urpmi.cfg differ: byte 2907,
line 135 is 40 175 }
***************************************
