[Yum] yum-arch changes the file name

seth vidal skvidal at linux.duke.edu
Sun Mar 5 01:11:30 UTC 2006


On Thu, 2006-03-02 at 10:33 +1100, cdl at dodo.com.au wrote:
> Not ony does yum-arch create compressed headers whereas headers created by
> yum not compressed, but yum-arch inserts a number into the filename - after
> the  package name. For example:
> 
> packages/cups-1.1.23-15.4.i386.rpm
> headers/cups-1-1.1.23-15.4.i386.hdr
> 
> Why? How do i stop this?
> (I could script to rename them)
> 
> regards
> Chris
> 
> Earlier I said:
> 
> Date: Wed, 1 Mar 2006 13:02:03 +1100
> From: "" <cdl at dodo.com.au>
> Subject: [Yum] Header generation (for read-only cache)
> To: yum at lists.dulug.duke.edu
> 
> I wish to have a read only yum cache that is accessed by many machines.
> 
> My yum repository is set to file:///var/cache/yum/updates-released/packages.
> 
> I populate the "packages directory using wget in mirror mode. I also get the
> "repodata" subdirectory.
> 
> Now I can use yum directly:
> 
> As intended "yum check-update" gets files such as repomd.xml from
> /var/cache/yum/updates-released/packages/repodata/ and places them in
> /var/cache/yum/updates-released/.
> 
> A "yum install kernel-2.6.15-xxx.i686" works well and produces a header in
> /var/cache/yum/updates-released/headers/.
> ("file" reports the header type as "data")
> 
> I wish to create all headers so that I can run in cache-only mode on
> NFS-mounted clients. So I use "/usr/share/yum-cli/pullheaders.py". This
> produces the headers in the correct location, but the file type is reported
> differently:
> headers/cups-1-1.1.23-15.4.i386.hdr: gzip compressed data, was
> ".newheaders/cups-1-1.1.23-15.4.", max compression
> 
> What is my mistake?
> 

yum-arch is onyl intended to make the .hdr files for yum 1.0.X and yum
2.0.X releases. It is not meant for use with yum 2.2.X and above.

If you want to pregenerate the hdr files that yum 2.2.X uses for caching
purposes you can either steal the routine we use to download them or
setup a simple script to process a lot of rpms at once.

it's just a matter of reading data out of the rpm from a specific byte
range, really easy, actually.

-sv





More information about the Yum mailing list