[Spacewalk-list] RHN Errata in Spacewalk

[Mullis, Josh (CCI - Atlanta)]

Andy,

Great work on this script.
It is a very nice addition to spacewalk...


I am having one problem and it's probably something I'm doing wrong:
When I'm running the script, I get a good bit of the "Hmmm... Package
Missing:" errors.

For example, when I'm cloning the rhel-x86_64-server-5 errata from RHN,
kpartx is one of the packages that it says it's missing.

I can search for kpartx in the package search page and I have a main
package called "kpartx", and it has all the versions inside.



It might be a database problem, but I figured I'd see if you've seen
this before.

Using:
- spacewalk v0.6
- rhn-clone-errata.py v0.4


Many thanks!
-Josh


On Wed, 2009-10-14 at 08:07 -0400, Andy Speagle wrote:

> On Wed, 2009-10-14 at 03:30 -0500, RLehmann at timocom.com wrote:
> > is there a way to push RHN Erratas into Spacewalk 0.6 on an RHEL 5
> > Server?
> 
> The best way to do this is to use the API to replicate the errata from
> RHN into your Spacewalk server.  I've proven the concept on my own
> Spacewalk 0.6 server with a python script ... and a few caveats.
> 
> 1) There doesn't seem to be a way to efficiently "merge" errata like you
> would in a commercial Satellite with access to official RHN errata.  
> 
> 2) Publishing an entire erratum into all channels on your Spacewalk
> server for which it is relevant can result in undesired behaviour.  For
> instance, if you have an erratum that effects both RHEL4 and RHEL5
> (assuming you are managing both flavours) the RPMs meant for RHEL5 are
> pushed into the RHEL4 channel as well, since it's just checking version.
> 
> The best way I've found to avoid this is to split each erratum into
> multiple errata, one for each RHEL version and arch, and append a
> version-arch code to the end of the advisory name, such as,
> RHSA-XXXX:YYYY:R5-64 (RHEL5 - x86_64).
> 
> 3) Subsequent cloning of these errata for publishing into separate
> "production" channels doesn't work as one would expect from the WebUI.
> Cloning works fantastically, however, the publishing step strips the
> RPMs from the cloned erratum and does not push the relevant RPMs into
> the destination "production" channel.  I have filed a bug on this one.
> 
> The only way I've found to do this properly is via the API, where
> "server.errata.clone()" does indeed push the packages to the destination
> channel and retains them in the cloned errata.
> 
> 4) There's no way via the v0.6 API to specify the "issue date" for an
> erratum.  This date is merely the date in which you created the erratum.
> This makes cloning errata into "production" a little more tricky, since
> I can't rely on "issue date" in my Spacewalk server.  I have submitted a
> bug/feature request for that one too.
> 
> ----
> 
> I hope that helps.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/spacewalk-list/attachments/20091017/401ff31f/attachment.htm>